Podcasts about BigQuery

Balazs Molnar, CEO and co-founder of Rabbit, chats with Kieron Allen about the evolving challenges of cloud cost management and how engineering teams have become central to tackling them. He explains why traditional FinOps tools fall short, how Rabbit dives below the surface to uncover hidden waste (especially in platforms like BigQuery) and why automation is essential for real savings.Optimizing Cloud with RabbitThe Big Themes:Cloud Costs Take Center Stage: Companies are no longer asking, "What can we build on the cloud?" They're now asking, "Why is this so expensive?" Rabbit's origin stems from this exact pivot: cloud costs spiraled out of control, catching businesses off guard. Despite robust migration to cloud environments like Google Cloud, companies found themselves ill-equipped to understand the hidden inefficiencies causing waste. Cloud spend can quickly balloon without the right oversight.The Cloud Buffet Problem: Balazs described cloud computing like a buffet: Engineers can take whatever they want, whenever they want. The cloud's flexibility is its strength but also its greatest risk. Unlike traditional on-prem setups that required hardware purchases and physical limits, cloud environments are boundless. Engineering teams now hold the wheel, yet they're typically not tasked to steer toward efficiency. This creates what Molnar calls a "FinOps trap": assuming finance can solve a problem that's fundamentally technical.Why Optimization Matters Now: Cloud vendors are still growing at impressive rates, but cracks are forming. Some businesses are exiting the cloud, not because they dislike the model — but because costs feel unmanageable. Molnar warns that in most cases, this isn't a cloud problem — it's an optimization problem. The promise of cloud was flexibility and scalability. But without proper tools, it becomes unpredictably expensive.The Big Quote: "We all know the news that cloud vendors are growing 30%+ on a year-over-year basis. But we also started to see cracks in the system where companies are actually deciding to move out of the cloud because it's too expensive to them. But the reality [is] it might not have to be that expensive. It's just not optimized."More from Balazs Molnar and Rabbit:Connect with Balazs on LinkedIn and check out more about Rabbit.* Sponsored podcast * 

In this episode, Lois Houston and Nikita Abraham dive into key components of Oracle GoldenGate 23ai with expert insights from Nick Wagner, Senior Director of Product Management.   They break down the Distribution Service, explaining how it moves trail files between environments, replaces the classic extract pump, and ensures secure data transfer. Nick also introduces Target Initiated Paths, a method for connecting less secure environments to more secure ones, and discusses how the Receiver Service simplifies monitoring and management. The episode wraps up with a look into Initial Load, covering different methods for syncing source and target databases without downtime.   Oracle GoldenGate 23ai: Fundamentals: https://mylearn.oracle.com/ou/course/oracle-goldengate-23ai-fundamentals/145884/237273 Oracle University Learning Community: https://education.oracle.com/ou-community LinkedIn: https://www.linkedin.com/showcase/oracle-university/ X: https://x.com/Oracle_Edu   Special thanks to Arijit Ghosh, David Wright, Kris-Ann Nansen, Radhika Banka, and the OU Studio Team for helping us create this episode. ----------------------------------------------------------------- Episode Transcript: 00:00 Welcome to the Oracle University Podcast, the first stop on your cloud journey. During this series of informative podcasts, we'll bring you foundational training on the most popular Oracle technologies. Let's get started! 00:25 Nikita: Welcome to the Oracle University Podcast! I'm Nikita Abraham, Team Lead of Editorial Services with Oracle University, and with me is Lois Houston, Director of Innovation Programs.  Lois: Hey there! Last week, we spoke about the Extract process and today we're going to spend time discussing the Distribution Path, Target Initiated Path, Receiver Server, and Initial Load. These are all critical components of the GoldenGate architecture, and understanding how they work together is essential for successful data replication. 00:58 Nikita: To help us navigate these topics, we've got Nick Wagner joining us again. Nick is a Senior Director of Product Management for Oracle GoldenGate. Hi Nick! Thanks for being with us today. To kick things off, can you tell us what the distribution service is and how it works? Nick: A distribution path is used when we need to send trail files between two different GoldenGate environments. The distribution service replaces the extract pump that was used in GoldenGate classic architecture. And so the distribution service will send the trail files as they're being created to that receiver service and it will write the trail files over on the target system. The distribution service works in a kind of a streaming fashion, so it's constantly pulling the trail files that the extract is creating to see if there's any new data. As soon as it sees new data, it'll packet it up and send it across the network to the receiver service. It can use a couple of different methods to do this. The most secure and recommended method is using a WebSocket secure connection or WSS. If you're going between a microservices and a classic architecture, you can actually tell the distribution service to send it using the classic architecture method. In that case, it's the OGG option when you're configuring the distribution service. There's also some unsecured methods that would send the trail files in plain text. The receiver service is then responsible for taking that data and rewriting it into the trail file on the target site. 02:23 Lois: Nick, what are some of the key features and responsibilities of the distribution service? Nick: It's responsible for command deployment. So any time that you're going to actually make a command to the distribution service, it gets handled there directly. It can handle multiple commands concurrently. It's going to dispatch trail files to one or more receiver servers so you can actually have a single distribution path, send trail files to multiple targets. It can provide some lightweight filtering so you can decide which tables get sent to the target system. And it also is integrated in with our data streams, our pub and subscribe model that we've added in GoldenGate 23ai. 03:01 Lois: Interesting. And are there any protocols to remember when using the distribution service? Nick: We always recommend a secure WebSocket. You also have proxy support for use within cloud environments. And then if you're going to a classic architecture GoldenGate, you would use the Oracle GoldenGate protocol. So in order to communicate with the distribution service and send it commands, you can communicate directly from any web browser, client software-- installation is not required-- or you can also do it through the admin client if necessary, but you can do it directly through browsers. 03:33 Nikita: Ok, let's move on to the target initiated path. Nick, what is it and what does it do essentially? Nick: This is used when you're communicating from a less secure environment to a more secure environment. Often, this requires going through some sort of DMZ. In these situations, a connection cannot be established from the less secure environment into the more secure environment. It actually needs to be established from the more secure environment out. And so if we need to replicate data into a more secure environment, we need to actually have the target GoldenGate environment initiate that connection so that it can be established.  And that's what a target-initiated path does. 04:12 Lois: And how do you set it up? Nick: It's pretty straightforward to set up. You actually don't even need to worry about it on the source side. You actually set it up and configure it from the target. The receiver service is responsible for receiving the trail file data and writing it to the local trail file. In this situation, we have a target-initiated path created. And so that receiver service is going to write the trail files locally and the replicat is going to apply that data into that target system. 04:37 Nikita: I also want to ask you about the Receiver service. What is it really? Nick: Receiver service is pretty straightforward. It's a centrally controlled service. It allows you to view the status of your distribution path and replaces target side collectors that were available in the classic architecture of GoldenGate. You can also get statistics about the receiver service directly from the web UI.  You can get detailed information about these paths by going into the receiver service and identifying information like network details, transfer protocols, how many bytes it's received, how many bytes it's sent out. If you need to issue commands from the admin client to the receiver service, you can use the info command to get details about it. Info all will tell you everything that's running. And you can see that your receiver service is up and running. 05:28 Are you working towards an Oracle Certification this year? Join us at one of our certification prep live events in the Oracle University Learning Community. Get insider tips from seasoned experts and learn from others who have already taken their certifications. Go to community.oracle.com/ou to jump-start your journey towards certification today! 05:53 Nikita: Welcome back. In the last section of today's episode, we'll cover what Initial Load is. Nick, can you break down the basics for us? Nick: So, the initial load is really used when you need to synchronize the source and target systems. Because GoldenGate is designed for 24/7 environments, we need to be able to do that initial load without taking downtime on the source. And so all the methods that we talk about do not require any downtime for that source database. 06:18 Lois: How do you do the initial load? Nick: So there's a couple of different ways to do the initial load. And it really depends on what your topology is. If I'm doing like-to-like replication in a homogeneous environment, we'll say Oracle-to-Oracle, the best options are to use something that's integrated with GoldenGate, some sort of precise instantiation method that does not require HandleCollisions. That's something like a database backup and restoring it to a specific SDN or CSN value using a Database Snapshot. Or in some cases, we can use Oracle Data Pump integration with GoldenGate. There are some less precise instantiation options, which do require HandleCollisions. We also have dissimilar initial load methods. And this is typically when you're going between heterogeneous environments. When my source and target databases don't match and there isn't any kind of fast unload or fast load utility that I could use between those two databases. In almost all cases, this does require HandleCollisions to be used. 07:16 Nikita: Got it. So, with so many options available, are there any advantages to using GoldenGate's own initial load method?  Nick: While some databases do have very good fast load and unload utilities, there are some advantages to using GoldenGate's own initial load method. One, it supports heterogeneous replication environments. So if I'm going from Postgres to Oracle, it'll do all the data type transformation, character set transformation for me. It doesn't require any downtime, if certain conditions are met.  It actually performs transformation as the data is loaded, too, as well as filtering. And so any transformation that you would be doing in your normal transaction log replication or CDC replication can also go through the same transformation for the initial load process. GoldenGate's initial load process does read directly from the source tables. And it fetches the data in arrays. It also uses parallel processing to speed up the replication. It does also handle activity on the source tables during the initial load process, so you do not need to worry about quiescing that source database. And a lot of the initial load methods directly built into GoldenGate support distributed application analytics targets, including things like Databricks, Snowflake, BigQuery. 08:28 Lois: And what about its limitations? Or to put it differently, when should users consider using different methods? Nick: So the first thing to consider is system proximity. We want to make sure that the two systems we're working with are close together. Or if not, how are we going to send the data across? One thing to keep in mind, when we do the initial load, the source database is not quiesced. So if it takes an hour to do the initial load or 10 hours, it really doesn't matter to GoldenGate. So that's something to keep in mind. Even though we talk about performance of this, the performance really isn't as critical as one might suspect. So the important thing about data system proximity is the proximity to the extract and replicat processes that are going to be pulling the data out and pushing it across. And then how much data is generated? Are we talking about a database that's just a couple of gigabytes? Or are we talking about a database that's hundreds of terabytes? Do we want to consider outage time? Would it be faster to take a little bit of outage and use some other method to move the data across? What kind of outage or downtime windows do we have for these environments? And then another consideration is disk space. As we're pulling the data out of that source database, we need to have somewhere to store it. And if we don't have enough disk space, we need to run to temporary space or to use multiple external drives to be able to support it. So these are all different considerations. 09:50 Nikita: I think we can wind up our episode with that. Thanks, Nick, for giving us your insights.  Lois: If you'd like to learn more about the topics we covered today, head over to mylearn.oracle.com and check out the Oracle GoldenGate 23ai: Fundamentals course. Nikita: In our next episode, Nick will take us through the Replicat process. Until then, this is Nikita Abraham… Lois: And, Lois Houston signing off! 10:14 That's all for this episode of the Oracle University Podcast. If you enjoyed listening, please click Subscribe to get all the latest episodes. We'd also love it if you would take a moment to rate and review us on your podcast app. See you again on the next episode of the Oracle University Podcast.

Dans ce nouvel épisode de "IA pas que la Data", on délaisse un peu les modèles pour plonger tête la première dans l'actualité brûlante du mois de mai. Pierre et Thomas débattent Vibe Coding, des nouvelles de Google, et surtout : jusqu'où l'IA va-t-elle changer le monde du travail ? Préparez-vous à une conversation passionnante et sans langue de bois !(00:00) - Introduction (01:40) - Tour d'horizon des activités : événements et retours d'expériences (02:20) - Le Vibe Coding et les événements à Montréal (05:25) - Generative AI Montréal et les Test Containers (06:09) - Google Next : agents et BigQuery (07:22) - Retour sur Cursor et les assistants au coding (10:35) - Les risques de la dette technique due au Vibe Coding (14:05) - Comparaison des assistants au coding : Codex et Claude Code (17:32) - Modèles de tarification et offre Google Ultra (21:36) - L'impact de l'IA sur l'industrie, la pub et le consulting (22:32) - VO3 et Flow chez Google : Text to Video et édition vidéo (24:21) - Microsoft, OpenAI et les chiffres (25:23) - L'émergence des "full stack startups" et le futur du dév (38:19) - La redéfinition des métiers et le rôle de l'humain (46:27) - Les enjeux sociétaux et politiques de l'IA Cet épisode vous a plu ? Dites-le nous avec 5 étoiles et partagez-le !

In this episode, Lois Houston and Nikita Abraham continue their deep dive into Oracle GoldenGate 23ai, focusing on its evolution and the extensive features it offers. They are joined once again by Nick Wagner, who provides valuable insights into the product's journey.   Nick talks about the various iterations of Oracle GoldenGate, highlighting the significant advancements from version 12c to the latest 23ai release. The discussion then shifts to the extensive new features in 23ai, including AI-related capabilities, UI enhancements, and database function integration.   Oracle GoldenGate 23ai: Fundamentals: https://mylearn.oracle.com/ou/course/oracle-goldengate-23ai-fundamentals/145884/237273 Oracle University Learning Community: https://education.oracle.com/ou-community LinkedIn: https://www.linkedin.com/showcase/oracle-university/ X: https://x.com/Oracle_Edu   Special thanks to Arijit Ghosh, David Wright, Kris-Ann Nansen, Radhika Banka, and the OU Studio Team for helping us create this episode.   -----------------------------------------------------------------   Episode Transcript: 00:00 Welcome to the Oracle University Podcast, the first stop on your cloud journey. During this series of informative podcasts, we'll bring you foundational training on the most popular Oracle technologies. Let's get started! 00:25 Lois: Hello and welcome to the Oracle University Podcast! I'm Lois Houston, Director of Innovation Programs with Oracle University, and with me is Nikita Abraham, Team Lead: Editorial Services.  Nikita: Hi everyone! Last week, we introduced Oracle GoldenGate and its capabilities, and also spoke about GoldenGate 23ai. In today's episode, we'll talk about the various iterations of Oracle GoldenGate since its inception. And we'll also take a look at some new features and the Oracle GoldenGate product family. 00:57 Lois: And we have Nick Wagner back with us. Nick is a Senior Director of Product Management for GoldenGate at Oracle. Hi Nick! I think the last time we had an Oracle University course was when Oracle GoldenGate 12c was out. I'm sure there's been a lot of advancements since then. Can you walk us through those? Nick: GoldenGate 12.3 introduced the microservices architecture. GoldenGate 18c introduced support for Oracle Autonomous Data Warehouse and Autonomous Transaction Processing Databases. In GoldenGate 19c, we added the ability to do cross endian remote capture for Oracle, making it easier to set up the GoldenGate OCI service to capture from environments like Solaris, Spark, and HP-UX and replicate into the Cloud. Also, GoldenGate 19c introduced a simpler process for upgrades and installation of GoldenGate where we released something called a unified build. This means that when you install GoldenGate for a particular database, you don't need to worry about the database version when you install GoldenGate. Prior to this, you would have to install a version-specific and database-specific version of GoldenGate. So this really simplified that whole process. In GoldenGate 23ai, which is where we are now, this really is a huge release.  02:16 Nikita: Yeah, we covered some of the distributed AI features and high availability environments in our last episode. But can you give us an overview of everything that's in the 23ai release? I know there's a lot to get into but maybe you could highlight just the major ones? Nick: Within the AI and streaming environments, we've got interoperability for database vector types, heterogeneous capture and apply as well. Again, this is not just replication between Oracle-to-Oracle vector or Postgres to Postgres vector, it is heterogeneous just like the rest of GoldenGate. The entire UI has been redesigned and optimized for high speed. And so we have a lot of customers that have dozens and dozens of extracts and replicats and processes running and it was taking a long time for the UI to refresh those and to show what's going on within those systems. So the UI has been optimized to be able to handle those environments much better. We now have the ability to call database functions directly from call map. And so when you do transformation with GoldenGate, we have about 50 or 60 built-in transformation routines for string conversion, arithmetic operation, date manipulation. But we never had the ability to directly call a database function. 03:28 Lois: And now we do? Nick: So now you can actually call that database function, database stored procedure, database package, return a value and that can be used for transformation within GoldenGate. We have integration with identity providers, being able to use token-based authentication and integrate in with things like Azure Active Directory and your other single sign-on for the GoldenGate product itself. Within Oracle 23ai, there's a number of new features. One of those cool features is something called lock-free reservation columns. So this allows you to have a row, a single row within a table and you can identify a column within that row that's like an inventory column. And you can have multiple different users and multiple different transactions all updating that column within that same exact row at that same time. So you no longer have row-level locking for these reservation columns. And it allows you to do things like shopping carts very easily. If I have 500 widgets to sell, I'm going to let any number of transactions come in and subtract from that inventory column. And then once it gets below a certain point, then I'll start enforcing that row-level locking. 04:43 Lois: That's really cool… Nick: The one key thing that I wanted to mention here is that because of the way that the lock-free reservations work, you can have multiple transactions open on the same row. This is only supported for Oracle to Oracle. You need to have that same lock-free reservation data type and availability on that target system if GoldenGate is going to replicate into it. 05:05 Nikita: Are there any new features related to the diagnosability and observability of GoldenGate?  Nick: We've improved the AWR reports in Oracle 23ai. There's now seven sections that are specific to Oracle GoldenGate to allow you to really go in and see exactly what the GoldenGate processes are doing and how they're behaving inside the database itself. And there's a Replication Performance Advisor package inside that database, and that's been integrated into the Web UI as well. So now you can actually get information out of the replication advisor package in Oracle directly from the UI without having to log into the database and try to run any database procedures to get it. We've also added the ability to support a per-PDB Extract.  So in the past, when GoldenGate would run on a multitenant database, a multitenant database in Oracle, all the redo data from any pluggable database gets sent to that one redo stream. And so you would have to configure GoldenGate at the container or root level and it would be able to access anything at any PDB. Now, there's better security and better performance by doing what we call per-PDB Extract. And this means that for a single pluggable database, I can have an extract that runs at that database level that's going to capture information just from that pluggable database. 06:22 Lois And what about non-Oracle environments, Nick? Nick: We've also enhanced the non-Oracle environments as well. For example, in Postgres, we've added support for precise instantiation using Postgres snapshots. This eliminates the need to handle collisions when you're doing Postgres to Postgres replication and initial instantiation. On the GoldenGate for big data side, we've renamed that product more aptly to distributed applications in analytics, which is really what it does, and we've added a whole bunch of new features here too. The ability to move data into Databricks, doing Google Pub/Sub delivery. We now have support for XAG within the GoldenGate for distributed applications and analytics. What that means is that now you can follow all of our MAA best practices for GoldenGate for Oracle, but it also works for the DAA product as well, meaning that if it's running on one node of a cluster and that node fails, it'll restart itself on another node in the cluster. We've also added the ability to deliver data to Redis, Google BigQuery, stage and merge functionality for better performance into the BigQuery product. And then we've added a completely new feature, and this is something called streaming data and apps and we're calling it AsyncAPI and CloudEvent data streaming. It's a long name, but what that means is that we now have the ability to publish changes from a GoldenGate trail file out to end users. And so this allows through the Web UI or through the REST API, you can now come into GoldenGate and through the distributed applications and analytics product, actually set up a subscription to a GoldenGate trail file. And so this allows us to push data into messaging environments, or you can simply subscribe to changes and it doesn't have to be the whole trail file, it can just be a subset. You can specify exactly which tables and you can put filters on that. You can also set up your topologies as well. So, it's a really cool feature that we've added here. 08:26 Nikita: Ok, you've given us a lot of updates about what GoldenGate can support. But can we also get some specifics? Nick: So as far as what we have, on the Oracle Database side, there's a ton of different Oracle databases we support, including the Autonomous Databases and all the different flavors of them, your Oracle Database Appliance, your Base Database Service within OCI, your of course, Standard and Enterprise Edition, as well as all the different flavors of Exadata, are all supported with GoldenGate. This is all for capture and delivery. And this is all versions as well. GoldenGate supports Oracle 23ai and below. We also have a ton of non-Oracle databases in different Cloud stores. On an non-Oracle side, we support everything from application-specific databases like FairCom DB, all the way to more advanced applications like Snowflake, which there's a vast user base for that. We also support a lot of different cloud stores and these again, are non-Oracle, nonrelational systems, or they can be relational databases. We also support a lot of big data platforms and this is part of the distributed applications and analytics side of things where you have the ability to replicate to different Apache environments, different Cloudera environments. We also support a number of open-source systems, including things like Apache Cassandra, MySQL Community Edition, a lot of different Postgres open source databases along with MariaDB. And then we have a bunch of streaming event products, NoSQL data stores, and even Oracle applications that we support. So there's absolutely a ton of different environments that GoldenGate supports. There are additional Oracle databases that we support and this includes the Oracle Metadata Service, as well as Oracle MySQL, including MySQL HeatWave. Oracle also has Oracle NoSQL Spatial and Graph and times 10 products, which again are all supported by GoldenGate. 10:23 Lois: Wow, that's a lot of information! Nick: One of the things that we didn't really cover was the different SaaS applications, which we've got like Cerner, Fusion Cloud, Hospitality, Retail, MICROS, Oracle Transportation, JD Edwards, Siebel, and on and on and on.  And again, because of the nature of GoldenGate, it's heterogeneous. Any source can talk to any target. And so it doesn't have to be, oh, I'm pulling from Oracle Fusion Cloud, that means I have to go to an Oracle Database on the target, not necessarily.  10:51 Lois: So, there's really a massive amount of flexibility built into the system.  11:00 Unlock the power of AI Vector Search with our new course and certification. Get more accurate search results, handle complex datasets easily, and supercharge your data-driven decisions. From now through May 15, 2025, we are waiving the certification exam fee (valued at $245). Visit mylearn.oracle.com to enroll. 11:26 Nikita: Welcome back! Now that we've gone through the base product, what other features or products are in the GoldenGate family itself, Nick? Nick: So we have quite a few. We've kind of touched already on GoldenGate for Oracle databases and non-Oracle databases. We also have something called GoldenGate for Mainframe, which right now is covered under the GoldenGate for non-Oracle, but there is a licensing difference there. So that's something to be aware of. We also have the OCI GoldenGate product. We are announcing and we have announced that OCI GoldenGate will also be made available as part of the Oracle Database@Azure and Oracle Database@ Google Cloud partnerships.  And then you'll be able to use that vendor's cloud credits to actually pay for the OCI GoldenGate product. One of the cool things about this is it will have full feature parity with OCI GoldenGate running in OCI. So all the same features, all the same sources and targets, all the same topologies be able to migrate data in and out of those clouds at will, just like you do with OCI GoldenGate today running in OCI.  We have Oracle GoldenGate Free.  This is a completely free edition of GoldenGate to use. It is limited on the number of platforms that it supports as far as sources and targets and the size of the database.  12:45 Lois: But it's a great way for developers to really experience GoldenGate without worrying about a license, right? What's next, Nick? Nick: We have GoldenGate for Distributed Applications and Analytics, which was formerly called GoldenGate for big data, and that allows us to do all the streaming. That's also where the GoldenGate AsyncAPI integration is done. So in order to publish the GoldenGate trail files or allow people to subscribe to them, it would be covered under the Oracle GoldenGate Distributed Applications and Analytics license. We also have OCI GoldenGate Marketplace, which allows you to run essentially the on-premises version of GoldenGate but within OCI. So a little bit more flexibility there. It also has a hub architecture. So if you need that 99.99% availability, you can get it within the OCI Marketplace environment. We have GoldenGate for Oracle Enterprise Manager Cloud Control, which used to be called Oracle Enterprise Manager. And this allows you to use Enterprise Manager Cloud Control to get all the statistics and details about GoldenGate. So all the reporting information, all the analytics, all the statistics, how fast GoldenGate is replicating, what's the lag, what's the performance of each of the processes, how much data am I sending across a network. All that's available within the plug-in. We also have Oracle GoldenGate Veridata. This is a nice utility and tool that allows you to compare two databases, whether or not GoldenGate is running between them and actually tell you, hey, these two systems are out of sync. And if they are out of sync, it actually allows you to repair the data too. 14:25 Nikita: That's really valuable…. Nick: And it does this comparison without locking the source or the target tables. The other really cool thing about Veridata is it does this while there's data in flight. So let's say that the GoldenGate lag is 15 or 20 seconds and I want to compare this table that has 10 million rows in it. The Veridata product will go out, run its comparison once. Once that comparison is done the first time, it's then going to have a list of rows that are potentially out of sync. Well, some of those rows could have been moved over or could have been modified during that 10 to 15 second window. And so the next time you run Veridata, it's actually going to go through. It's going to check just those rows that were potentially out of sync to see if they're really out of sync or not. And if it comes back and says, hey, out of those potential rows, there's two out of sync, it'll actually produce a script that allows you to resynchronize those systems and repair them. So it's a very cool product.  15:19 Nikita: What about GoldenGate Stream Analytics? I know you mentioned it in the last episode, but in the context of this discussion, can you tell us a little more about it?  Nick: This is the ability to essentially stream data from a GoldenGate trail file, and they do a real time analytics on it. And also things like geofencing or real-time series analysis of it.  15:40 Lois: Could you give us an example of this? Nick: If I'm working in tracking stock market information and stocks, it's not really that important on how much or how far down a stock goes. What's really important is how quickly did that stock rise or how quickly did that stock fall. And that's something that GoldenGate Stream Analytics product can do. Another thing that it's very valuable for is the geofencing. I can have an application on my phone and I can track where the user is based on that application and all that information goes into a database. I can then use the geofencing tool to say that, hey, if one of those users on that app gets within a certain distance of one of my brick-and-mortar stores, I can actually send them a push notification to say, hey, come on in and you can order your favorite drink just by clicking Yes, and we'll have it ready for you. And so there's a lot of things that you can do there to help upsell your customers and to get more revenue just through GoldenGate itself. And then we also have a GoldenGate Migration Utility, which allows customers to migrate from the classic architecture into the microservices architecture. 16:44 Nikita: Thanks Nick for that comprehensive overview.  Lois: In our next episode, we'll have Nick back with us to talk about commonly used terminology and the GoldenGate architecture. And if you want to learn more about what we discussed today, visit mylearn.oracle.com and take a look at the Oracle GoldenGate 23ai Fundamentals course. Until next time, this is Lois Houston… Nikita: And Nikita Abraham, signing off! 17:10 That's all for this episode of the Oracle University Podcast. If you enjoyed listening, please click Subscribe to get all the latest episodes. We'd also love it if you would take a moment to rate and review us on your podcast app. See you again on the next episode of the Oracle University Podcast.

Full show notes, transcript and AI chatbot - https://bit.ly/42SW2yG Watch on YouTube - https://youtu.be/08gLvHthNuw 00:00:00 – Introduction 00:05:37 – ChatGPT shopping and e-commerce 00:08:50 – AI impact on data tracking 00:12:28 – Natural language data analysis tool 00:16:18 – Discovery Catalog for Cloud Storage 00:19:11 – Data accessibility in BigQuery 00:25:23 – AI-generated insights in GA4 00:27:56 – Cloud Code features in IDE 00:30:36 – Data robustness in BigQuery 00:35:45 – Data layer responsibilities in marketing 00:39:47 – Importance of monitoring data integrity 00:42:51 – Documentation and data governance 00:46:56 – Derelict data warehouse issues 00:49:12 – Data governance and documentation 00:54:26 – Principles of least privilege 00:58:03 – Infrastructure as code with Terraform 01:00:04 – Audience participation ideas ----- Episode Summary: In this episode of The Measure Pod, Dara and Matt dive into a whole mix of marketing analytics news and musings. From new BigQuery features and GA4 updates to ChatGPT integrations and data quality best practices, they cover what's actually worth knowing. It's a loose but lively chat full of useful insights, personal takes, and a healthy dose of “pinch of salt” news. Expect talk of snooker, rogue GA4 permissions, and why naming conventions are the hill we'll die on. ----- About The Measure Pod: The Measure Pod is your go-to fortnightly podcast hosted by seasoned analytics pros. Join Dara Fitzgerald (Co-Founder at Measurelab) & Matthew Hooson (Head of Engineering at Measurelab) as they dive into the world of data, analytics and measurement—with a side of fun. ----- If you liked this episode, don't forget to subscribe to The Measure Pod on your favourite podcast platform and leave us a review. Let's make sense of the analytics industry together! The post #120 Marketing analytics updates and a BigQuery health check appeared first on Measurelab.

In this episode of Game Time Tech, Robert Kramer and Melody Brue, VP and Principal Analysts at Moor Insights & Strategy, dive into the intersection of sports and technology. Explore how cutting-edge technologies like AI, data analytics, and personalized fan experiences are transforming Major League Baseball, The Masters Golf Tournament, and the Intuit Dome. Highlights include: MLB's Data Evolution with Google Cloud: Data and AI are enhancing fan engagement, team strategies, and broadcasting through platforms like Google Cloud's BigQuery. Masters Golf Tournament's AI Innovations: IBM's Generative AI is powering predictive insights for fans with features like "Every Shot" and "Every Hole" in The Masters app. Inside the Intuit Dome: A look at advanced fan experience technologies, including facial recognition for entry, autonomous stores, and real-time analytics, powered by Teradata. Mercedes-Benz's Cutting-Edge In-Car Experience: Technology is transforming connectivity, from live sports streaming to Zoom calls integrated directly into vehicles. Fan Behavior and Smart Stadiums: How data-driven technologies are shaping stadium interactions, from personalized fan experiences to autonomous retail systems.

This week, Frank sat down with Dr. Jacob Leverich—Stanford PhD, cofounder of Observe, and a veteran of the Google MapReduce team and Splunk. Jacob's journey, from tinkering with video game code as a kid, to innovating at the cutting edge of distributed systems and energy efficiency, is as inspiring as it is informative.Key TakeawaysEarly Tech Roots: Hear how curiosity with QBasic and classic PCs (think IBM PCXT and Commodore) put Jacob on a path to high-impact data engineering.MapReduce, Dremel, & the Rise of Big Data: Jacob pulls back the curtain on working with some of the most influential data processing tools at Google and how these systems shifted the entire data landscape (hello, BigQuery!).Building Efficient Systems: It's not just about scale—energy efficiency and performance optimization are the unsung heroes of today's data infrastructure. Jacob explains why making things “just work” isn't enough anymore.The Realities of Ops & Observability: Remember the days of grepping logs at 2AM? There's a better way. Jacob shares how platforms like Observe help teams consolidate, visualize, and act on operational data—turning chaos into actionable insight.Bridging Data & Ops: The lines between data observability and traditional ops are blurring, and Jacob's unique experience shows how best practices from data warehousing are finally making ops smoother (and less sleepless).Power Concerns & the Future: As data grows, so does energy consumption in data centers. Find out why optimization isn't just good for performance—it's key to sustainability.Timestamps00:00 Interview with Jacob Levrich05:59 Journey into Game Programming06:43 "Pursuing Fast Video Game Code"10:23 Data Processing and Power Efficiency16:11 Snowflake's Transformative Database Approach19:18 Journey to Data Management Industry21:37 Data Products: Solving Core Challenges27:07 Early Web Log Analysis Techniques28:57 Consolidating Data for Efficiency33:23 Specialized Tools and Context Switching35:43 Unique Dual-Expertise in Tech38:58 User-Centric Business Strategies42:13 IP Data Analysis in Cloud47:23 Electricity Transport Upsets Local Farms48:25 Shift to Parallel Computing52:10 Hardware Specialization & Software Optimization57:32 "Stay Data Driven"

Full show notes, transcript and AI chatbot - https://bit.ly/3Gg5HHZ Watch on YouTube - https://youtu.be/dcZhmVY_Bl0 00:00:00 - New co-host introduction. 00:04:01 - Google Next 25 conference highlights. 00:08:10 - CapEx spend on cloud and AI. 00:12:37 - Cross-cloud collaboration and flexibility. 00:15:40 - Gemini's integration in Firebase. 00:21:01 - Autonomous data AI platform. 00:25:10 - Data tools and data quality. 00:27:01 - Data quality challenges and solutions. 00:30:15 - Building with good foundations. 00:36:08 - Unstructured data in AI platforms. 00:40:10 - BigQuery as enterprise advantage. 00:42:56 - BigQuery vector search capabilities. 00:48:11 - Multi-agent systems and autonomy. 00:51:20 - Importance of robust data. 00:54:06 - BigQuery and unstructured data. 00:58:05 - Reducing repetitive work through automation. ----- Episode Summary: In this episode of The Measure Pod, Dara and Matthew take the reins and dive into the biggest takeaways from Google Cloud Next 2025. From shiny new features to subtle shifts in direction, they cover the bits that matter—what's exciting, what's useful, and what might actually change the way we work. Plenty of ground covered. Plenty of thoughts shared. And just the beginning of what's to come. ----- About The Measure Pod: The Measure Pod is your go-to fortnightly podcast hosted by seasoned analytics pros. Join Dara Fitzgerald (Co-Founder at Measurelab) & Matthew Hooson (Head of Engineering at Measurelab) as they dive into the world of data, analytics and measurement—with a side of fun. ----- If you liked this episode, don't forget to subscribe to The Measure Pod on your favourite podcast platform and leave us a review. Let's make sense of the analytics industry together! The post #119 Google Cloud Next 25 roundup appeared first on Measurelab.

In this episode we discuss the latest and greatest announcements from the Google Next 2025 conference with Simon Pane (Oracle ACE and Google Cloud Champion), Nelson Calero (Oracle Ace Director) and Jeff Deverter (Pythian Field CTO). We go over Oracle partnership updates, BigQuery updates, AlloyDb updates and of course, AI announcements!

Welcome to episode 298 of The Cloud Pod – where the forecast is always cloudy! Justin, Matthew and Ryan are in the house (and still very much missing Jonathan) to bring you a  jam packed show this week, with news from Beijing to Virginia! Did you know Virginia was in the US? Amazon definitely wants you to know that.  We've got updates from BigQuery Git Support and their new collab tools, plus all the AI updates you were hoping you'd miss. Tune in now!  Titles we almost went with this week: The Cloud Pod now Recorded from Planet Earth Wait Java still exists? When will java just be coffee and not software Cloudflare Makes AI beat Mazes Replacing native mobile things with mobile web apps won't fix your problems AWS Turn your security over to the bots The Cloud Pod is lost in the AI labyrinth  AI security agents to secure the AI… wait recursion Durable + Stateless.. I don't know if you know what those words means Click ops expands to our phones yay! The Cloud Pod is now a data analyst  Gitops come to bigquery A big thanks to this week's sponsor: We're sponsorless! Want to get your brand, company, or service in front of a very enthusiastic group of cloud news seekers? You've come to the right place! Send us an email or hit us up on our slack channel for more info.  AI Is Going Great – Or How ML Makes All Its Money   00:46 Manus, a New AI Agent From China is Going Viral—And Raising Big Questions   Manus is being described as “the first true autonomous AI agent” from China, capable of completing weeks of professional work in hours. Developed by a team called Butterfly Effect with offices in Beijing and Wuhan, Manus functions as a truly autonomous agent that independently analyzes, plans, and executes complex tasks.  The system uses a multi-agent architecture powered by several distinct AI models, including Anthropic’s Claude 3.5 Sonnet and fine-tuned versions of

Thu, 27 Mar 2025 23:00:00 +0000 https://mydata.podigee.io/263-new-episode 62e289047630ff510f640b895cd2984d Wie revolutioniert man Private Equity mit Daten? Wie trifft man bessere Investmententscheidungen? Und warum ist Datenkultur gerade in dieser Branche so wichtig? Darum geht es in der neuen Folge von MY DATA IS BETTER THAN YOURS, in der Host Jonas Rashedi mit Daniel Lebe spricht. Dieser verantwortet als Business Intelligence Developer bei FSN Capital die Themen Datenanalyse, Prozessoptimierung und Business Intelligence. Im Gespräch der beiden Data-Enthusiasten geht es zunächst um den digitalen Wandel bei FSN Capital. Das Unternehmen hat in den letzten drei Jahren eine komplette Datentransformation durchlaufen. Das Ziel: Bessere Deals durch bessere Daten! Daniel ist Teil eines sechsköpfigen Teams, welches sich um die digitale Transformation der FSN Deal Prozesse und um die akquirierten Portfolio Unternehmen kümmert. Diese Teams setzen sich aus ausgebildeten Data Scientists, einem Data Engineer und weiteren Spezialisten zusammen. Im Private-Equity Mid Market Bereich ein schlagkräftiges Team. Der Aufbau einer modernen Dateninfrastruktur war dabei die größte Herausforderung. Für Daniel liegt der Fokus darauf, die verschiedenen Stakeholder mit ihren unterschiedlichen Bedürfnissen zusammenzubringen und datenbasierte Entscheidungsgrundlagen zu schaffen. Ein besonderer Schwerpunkt liegt auf der Analyse erfolgreicher Deals. Daniel erzählt von seiner aktuellen Aufgabe, vergangene Investments zu analysieren, um daraus für zukünftige Entscheidungen zu lernen. Dabei ist ihm wichtig, möglichst einfach anzufangen und schrittweise die richtigen Fragen zu stellen. Bei dem Aufbau der Dateninfrastruktur setzt FSN Capital auf moderne Tools. Zum Beispiel wird Power BI für Visualisierungen genutzt und BigQuery als Data Warehouse. Zum Schluss spricht Daniel noch über seine persönliche Datenreise: Wie er selbst Ziele messbar macht und warum der Film "Edge of Tomorrow" sein Data Game am besten beschreibt - manchmal braucht es mehrere Iterationen und auch Rückschläge, um am Ende zum Erfolg zu kommen. MY DATA IS BETTER THAN YOURS ist ein Projekt von BETTER THAN YOURS, der Marke für richtig gute Podcasts. Zum LinkedIn-Profil von Daniel: https://de.linkedin.com/in/daniel-lebe-a75011155 Zur Webseite von FSN Capital: https://www.fsncapital.com/en/ Zu allen wichtigen Links rund um Jonas und den Podcast: https://linktr.ee/jonas.rashedi Zeitstempel: 00:00:00 Intro und Begrüßung 00:02:05 FSN Capital und Private Equity 00:05:10 Das Digital Team 00:06:57 Datenbasierte Investments 00:11:53 Herausforderungen der Datenanalyse 00:14:06 Moderne Dateninfrastruktur 00:22:47 Spannende Use Cases 00:26:53 Effektivität und Effizienz 00:28:55 Diversität im Team 00:33:51 Persönliche Datenziele 00:35:52 Daniels Data Game full no Datapodcast,Digitale Transformation,Datenanalyse,Investment,Power BI,Use Cases,Stakeholder Management,Analyse Podcast,Datenstrategie,Datenvisualisierung

Send us a textLet's demystify the magic behind streamlined customer success operations. In this episode of the Customer Success Playbook podcast, Kevin Metzger sits down with Gilad Shriki from Scope to unpack their strategic integration of FunnelStory. They dive into privacy-first data management, lightning-fast time-to-value, and how AI is reshaping how teams interact with data. Plus, find out why Gilad believes FunnelStory might just be the one platform to rule them all.Detailed Description with Business Insights: In this engaging episode of the Customer Success Playbook, Kevin Metzger interviews Gilad Shriki, Head of Customer Experience at Scope, who offers a real-world case study of successfully implementing FunnelStory. With Roman Trebon off this week, Kevin navigates a thoughtful conversation that brings valuable technical and strategic takeaways to customer success leaders.Gilad breaks down how Scope maintains data privacy by leveraging a custom anonymization layer before syncing anonymized data into BigQuery. From there, FunnelStory becomes the centerpiece of their CS tech stack, tightly integrated with HubSpot and Segment. The result? A seamless, compliant, and highly performant system that delivers actionable insights with minimal setup.The discussion peels back the curtain on modern data stack integrations, emphasizing the importance of time-to-value and the benefits of designing for automation-first customer success platforms. Gilad candidly explains how FunnelStory outperformed expectations by offering an intuitive plug-and-play experience and how its engineering team's responsiveness created a frictionless implementation.Most notably, Gilad envisions FunnelStory not just as a visibility tool but as a centralized hub for both automation and human interaction. His goal? A single pane of glass where CSMs manage sentiment, risk, and engagement—without needing to bolt on other platforms like Gainsight.If you're scaling a CS org or rethinking your tech stack, this episode is your playbook for staying lean without sacrificing power. Tune in and learn how a privacy-first, AI-powered, integrated system can revolutionize how you scale customer success.Now you can interact with us directly by leaving a voice message at https://www.speakpipe.com/CustomerSuccessPlaybookPlease Like, Comment, Share and Subscribe. You can also find the CS Playbook Podcast:YouTube - @CustomerSuccessPlaybookPodcastTwitter - @CS_PlaybookYou can find Kevin at:Metzgerbusiness.com - Kevin's person web siteKevin Metzger on Linked In.You can find Roman at:Roman Trebon on Linked In.

Send us a textIn this engaging episode of the Customer Success Playbook Podcast, host Kevin Metzger sits down with Gilad Shriki from The Scope to explore how FunnelStory is transforming customer success operations. With seamless integration capabilities and a robust automation-first approach, FunnelStory is setting a new standard for customer success platforms.Gilad shares insights into how his team successfully integrated FunnelStory with BigQuery, HubSpot, and Segment, all while maintaining strict data privacy protocols. He also discusses how AI-driven automation is enhancing customer sentiment analysis and churn prediction, giving CS teams an edge in proactive engagement.Is Funnel Story truly a one-stop shop for customer success? Can businesses of all sizes leverage its automation without sacrificing human interaction? Listen in as Gilad provides a firsthand account of his experience and why he believes FunnelStory is reshaping the future of customer success management.Detailed Episode Insights:Seamless Integration: How The Scope connected FunnelStory with their existing data stack while maintaining PII privacy.Automation at the Core: Why starting with automation before layering in human interaction changes the game for CS teams.AI-Powered Efficiency: How FunnelStory is accelerating time-to-value and making predictive insights more accessible.Scalability & Growth: Can FunnelStory support businesses up to $500M in revenue? Gilad shares his perspective.The Future of CS Tech: What's next for AI-powered customer success platforms?Now you can interact with us directly by leaving a voice message at https://www.speakpipe.com/CustomerSuccessPlaybookPlease Like, Comment, Share and Subscribe. You can also find the CS Playbook Podcast:YouTube - @CustomerSuccessPlaybookPodcastTwitter - @CS_PlaybookYou can find Kevin at:Metzgerbusiness.com - Kevin's person web siteKevin Metzger on Linked In.You can find Roman at:Roman Trebon on Linked In.

Taha Bel Khayate est Lead Analytics Engineer chez Brevo, la plateforme de marketing automation qui permet notamment d'orchestrer ses campagnes d'emailing ou de SMS. La scaleup a acquis le statut de “centaure” après avoir dépassé les 100 millions d'euros de revenus annuels. On revient sur l'un des plus gros challenges de l'équipe Analytics Engineering.

Topics covered in this episode: LLM Catcher On PyPI Quarantine process RESPX Unpacking kwargs with custom objects Extras Joke Watch on YouTube About the show Sponsored by us! Support our work through: Our courses at Talk Python Training The Complete pytest Course Patreon Supporters Connect with the hosts Michael: @mkennedy@fosstodon.org / @mkennedy.codes (bsky) Brian: @brianokken@fosstodon.org / @brianokken.bsky.social Show: @pythonbytes@fosstodon.org / @pythonbytes.fm (bsky) Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Monday at 10am PT. Older video versions available there too. Finally, if you want an artisanal, hand-crafted digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Michael #1: LLM Catcher via Pat Decker Large language model diagnostics for python applications and FastAPI applications . Features Exception diagnosis using LLMs (Ollama or OpenAI) Support for local LLMs through Ollama OpenAI integration for cloud-based models Multiple error handling approaches: Function decorators for automatic diagnosis Try/except blocks for manual control Global exception handler for unhandled errors from imported modules Both synchronous and asynchronous APIs Flexible configuration through environment variables or config file Brian #2: On PyPI Quarantine process Mike Fiedler Project Lifecycle Status - Quarantine in his "Safety & Security Engineer: First Year in Review post” Some more info now in Project Quarantine Reports of malware in a project kick things off Admins can now place a project in quarantine, allowing it to be unavailable for install, but still around for analysis. New process allows for packages to go back to normal if the report is false. However Since August, the Quarantine feature has been in use, with PyPI Admins marking ~140 reported projects as Quarantined. Of these, only a single project has exited Quarantine, others have been removed. Michael #3: RESPX Mock HTTPX with awesome request patterns and response side effects A simple, yet powerful, utility for mocking out the HTTPX, and HTTP Core, libraries. Start by patching HTTPX, using respx.mock, then add request routes to mock responses. For a neater pytest experience, RESPX includes a respx_mock fixture Brian #4: Unpacking kwargs with custom objects Rodrigo A class needs to have a keys() method that returns an iterable. a __getitem__() method for lookup Then double splat ** works on objects of that type. Extras Brian: A surprising thing about PyPI's BigQuery data - Hugovk Top PyPI Packages (and therefore also Top pytest Plugins) uses a BigQuery dataset Has grabbed 30-day data of 4,000, then 5,000, then 8,000 packages. Turns out 531,022 packages (amount returned when limit set to a million) is the same cost. So…. hoping future updates to these “Top …” pages will have way more data. Also, was planning on recording a Test & Code episode on pytest-cov today, but haven't yet. Hopefully at least a couple of new episodes this week. Finally updated pythontest.com with BlueSky links on home page and contact page. Michael: Follow up from Owen (uv-secure): Thanks for the multiple shout outs! uv-secure just uses the PyPi json API at present to query package vulnerabilities (same as default source for pip audit). I do smash it asynchronously for all dependencies at once... but it still takes a few seconds. Joke: Bugs hide from the light!

Juliette Duizabo est Head of Data chez Agorapulse, la startup qui propose un outil de gestion des réseaux sociaux qui a levé plus de 16 millions d'euros. Avant, Juliette était déjà Head of Data chez Ovrsea.

Send us a textWeb Crawler DesignsCan a simple idea like building a web crawler teach you the intricacies of system design? Join me, Ben Kitchell, as we uncover this fascinating intersection. Returning from a brief pause, I'm eager to guide you through the essential building blocks of a web crawler, from queuing seed URLs to parsing new links autonomously. These basic functionalities are your gateway to creating a minimum viable product or acing that system design interview. You'll gain insights into potential extensions like scheduled crawling and page prioritization, ensuring a strong foundation for tackling real-world challenges.Managing a billion URLs a month is no small feat, and scaling such a system requires meticulous planning. We'll break down the daunting numbers into digestible pieces, exploring how to efficiently store six petabytes of data annually. By examining different database models, you'll learn how to handle URLs, track visit timestamps, and keep data searchable. The focus is on creating a robust system that not only scales but does so in a way that meets evolving demands without compromising on performance.Navigating the complexities of designing a web crawler means making critical decisions about data storage and system architecture. We'll weigh the benefits of using cloud storage solutions like AWS S3 and Azure Blob Storage against maintaining dedicated servers. Discover the role of REST APIs in seamless user and service interactions, and explore search functionalities using Cassandra, Amazon Athena, or Google's BigQuery. Flexibility and foresight are key as we build systems that adapt to future needs. Thank you for your continued support—let's keep learning and growing on this exciting system design journey together.Support the showDedicated to the memory of Crystal Rose.Email me at LearnSystemDesignPod@gmail.comJoin the free Discord Consider supporting us on PatreonSpecial thanks to Aimless Orbiter for the wonderful music.Please consider giving us a rating on ITunes or wherever you listen to new episodes.

Google Cloud's Innovation and GrowthThe Big Themes:Google Cloud's record growth and market positioning: In 2024, Google Cloud experienced five consecutive quarters of accelerating growth, including a remarkable 35% growth in Q3, up from 29% in Q2. Kurian attributes this success to the company's ability to listen to customers, innovate with products that meet their evolving needs, and strategically invest in a strong go-to-market organization.AI cost reduction and efficiency: Kurian comments on Google Cloud's efforts to significantly reduce the cost of AI models. Through improved software stack capabilities and optimizations, Google has decreased the cost of AI by more than 10x in just six months. Reducing latency, improving response accuracy, and utilizing distillation (e.g., making models run on smaller devices like phones) have contributed to lowering operational costs while increasing model efficiency. This approach has resulted in a 15-17x growth in model usage in just five months.The evolving role of cloud in business transformation: Kurian notes a fundamental shift in how businesses view cloud computing. Initially seen as a way to reduce costs, cloud is now seen as a tool for driving business transformation. AI, analytics, and security capabilities are helping organizations speed up decision-making, optimize logistics, and gain competitive advantages. Kurian believes that the next wave of cloud adoption will focus more on enabling new business models, products, and markets rather than just reducing IT costs.The Big Quote: “We tend to look ahead by listening to customers and understanding their needs, and create in a disciplined way, new product offerings. If you look a the last five years, we've introduced enough steady cadence. First, we started with infrastructure, then we added databases to it. We used our strength with BigQuery to build out an analytics portfolio. We were one of the earliest to say . . . we should not only provide [customers] a secure cloud, but we should also build a security product portfolio. Every one of those has driven diversification of our revenue stream."

In this Checkout episode, we sit down with Jethro Marks, co-founder of The Nile, to uncover personal insights behind this pioneering ecom giant. Jethro shares his thoughts on disruptive platforms like Temu, his admiration for the logistics mastery of Dan Murphy's and the critical role Google's BigQuery is playing in powering The Nile. He also reflects on how balancing innovation with consistency has fed into the brand's long-term success amidst the ever-changing ecom landscape.Check out our full-length interview with Jethro Marks here:How Jethro Marks is Transforming The Nile into a Leading Aussie Online Bookstore | #454 This episode was brought to you by:Deliver In PersonShopify PlusAbout your guest:Jethro Marks is the Co-Founder and CEO of The Nile, one of Australia's pioneering pure-play online retailers. With over 15 years of experience in eCommerce, Jethro has been there since the start with co-founder Mark Taylor, taking the enterprise from a living room with two guys and a computer to a global operation across Australia, New Zealand, the US, and UK, offering over 40 million products. A former Director of NORA, he is also a Non-Executive Director at DroneShield (ASX: DRO).About your host:Nathan Bush is the host of the Add To Cart podcast and a leading ecommerce transformation consultant. He has led eCommerce for businesses with revenue $100m+ and has been recognised as one of Australia's Top 50 People in eCommerce four years in a row. You can contact Nathan on LinkedIn, Twitter or via email.Please contact us if you: Want to come on board as an Add To Cart sponsor Are interested in joining Add To Cart as a co-host Have any feedback or suggestions on how to make Add To Cart betterEmail hello@addtocart.com.au We look forward to hearing from you! Hosted on Acast. See acast.com/privacy for more information.

Bayer's Data Evolution with AlloyDBThe Big Themes:Data complexity and intelligent agriculture: Bayer Crop Science is addressing agriculture's complex data challenges. The company integrates data such as satellite imagery, weather conditions, soil data, and IoT device inputs, to drive innovation in seed development and farming practices. By leveraging cloud technologies like AlloyDB, Bayer's teams can support the future of farming, despite challenges posed by climate change and rising global food demand.Integrating BigQuery for comprehensive analytics: To further enhance its data-driven insights, Bayer integrates Google BigQuery alongside AlloyDB for extensive data analysis. BigQuery serves as the central analytics warehouse, receiving billions of phenotypic data points for in-depth modeling and decision-making. During harvest season, Bayer can quickly access and analyze comprehensive datasets, enabling better decisions across production and supply chains.Harvest season demands and system resilience: During harvest season, Bayer Crop Science faces intense pressure as high volumes of data flow in, requiring real-time analysis and decision-making. The peak demand period sees a sharp increase in read and write operations, making it essential for Bayer's data system to function seamlessly. AlloyDB played a crucial role in handling these spikes by providing low-latency data processing and high availability.The Big Quote: “Climate change is a new challenge. You see some of these forecasts coming out of academia that yields will go down by 30% — that will arrest this great trend that we've seen continually increasing over the last 100 years. We need to solve for that, and that's going to take new types of data and new approaches and these types of things."

Jack Chambers Ward sits down with web data analyst Marco Giordano to discuss how to get the most out of your web data. Together, they delve into the intricacies of web analytics, covering topics such as incorporating crawl data, understanding GA4, leveraging BigQuery, and effectively communicating data insights to clients. Jack and Marco also discuss the importance of combining technical SEO expertise with business acumen and provide valuable tips for anyone looking to maximise the impact of their web data. Sponsors AlsoAsked - Track search intent shift over time using the Timeline feature SE Ranking - Track AI Overviews for your clients using SE Ranking Links to follow Marco: Follow Marco on LinkedIn Check out SEOTistics Subscribe to Marco's newsletter Links to resources/articles: ⁠Free Google Cloud BigQuery training⁠ Analytics for SEO course ⁠Analytics for SEO ebook The Gray Dot keyword matrix template Chapters 00:00 Highlight reel 00:36 Welcome to Search with Candour 00:56 Introducing Marco Giordano 01:55 Sponsors 04:49 Marco Giordano's Insights on Web Data 06:12 Common Mistakes in Web Data Management 10:22 Balancing Hard and Soft Skills in SEO 19:00 Importance of Storing and Analysing Web Data 29:46 Combining Data Sources for Better Insights 37:17 Global Marketing Differences 39:26 The Role of Crawl Data and Log Files 42:05 Integrating Technical and Content Strategies 49:16 The Importance of Hybrid Skills in SEO 52:49 Evolving Job Titles and Industry Roles 54:27 Combining SEO with Other Channels 01:03:15 The Future of SEO and Analytics 01:09:17 Conclusion and Future Episodes

لو خطر في بالك قبل كده ليه عندنا كل قواعد البيانات دي, و ليه فيه منهم انواع مختلفة DBMS, NOSQL و غيرهم, طيب الناس اللي بتشتغل على الحاجات دي ايه التحديات اللي بيواجهوها, و ايه التخصص ده و ايه المتطلبات بتاعته. Ahmed Ayad is a SQL Engineer by trade, a database guy by education and training, and data dude by passion. I am currently an Engineering Director of the Managed Storage and Workload Management team in Google #BigQuery, building the best large scale enterprise data warehouse on the planet. My team owns the core parts of BigQuery involved in managing user data, metadata catalog, streaming and batch ingestion, replication, resource management and placement, physical sharding, and structured lake analytics. Over the years we have: - Grew data under management by several orders of magnitude. - Grew BigQuery's global footprint to more than 20+ regions and counting. - Enabled the hyper scaling of data analytics for a Who's Who list of Fortune 500 users, both Enterprise and Cloud-native. I am passionate about building cool technologies at scale, and the effective teams that create them. Things I did in previous professional lives: - I have shipped components in SQL Server product since SQL Server 2008. Worked on the Performance Data Collector, Policy Based Management, AlwaysOn, The Utility Control Point, SQL Azure stack from the backend to the middle-tier and Portal, SQL Server Agent, SQL Server Optimizer, and SQL Server Management Tools. - Did Database research in the areas of Data Mining, Query Optimization, and Data Streaming.

I am excited to bring you an insightful conversation with Russell Efird, Head of North American Partnerships at Quantum Metric, recorded live from Google Cloud's Marketplace Exchange! Russell dives into how Quantum Metric, a digital analytics experience platform, leverages the power of Google Cloud technologies like BigQuery and Gen AI to create seamless, high-performing digital journeys that resonate with C-level leaders and drive real business outcomes. Russell shares invaluable insights into the evolving enterprise buying landscape and the importance of aligning SaaS solutions to meet the needs of key decision-makers, from Chief Digital Officers to Heads of E-commerce. He highlights Quantum Metric's strategy of building “value networks” by collaborating with Google and other ISVs, enhancing the customer experience and accelerating business impact through innovative partnerships. Packed with practical strategies for growth, marketplace success, and ecosystem collaboration, this episode of The Ultimate Guide to Partnering is a must-watch for anyone invested in partnerships or digital analytics. Tune in for Russell's expert advice on building a future-focused partner strategy and driving growth through meaningful, multi-partner collaborations!

In this episode of SEO Cash Flow, it's me, Olga Zarr, teaming up with Myriam Jessier to tackle BigQuery for SEOs. We're diving into how you can pull more insights out of Google Search Console data without turning into a data scientist. Myriam's going all-in on learning BigQuery, while I'm sticking to my minimalist, ADHD-friendly approach—keeping it simple, powerful, and quick. We chat about why BigQuery isn't as scary as it seems and how it can give you way more control over your data, letting you see past the usual Google limits. This is for SEOs who want that edge without a ton of fuss or coding. If you've been wanting to get into BigQuery but didn't know where to start, this episode is your roadmap. Follow Myriam Jessier:

Welcome to episode 279 of The Cloud Pod, where the forecast is always cloudy! This week Justin, Jonathan and Matthew are your guide through the Cloud. We're talking about everything from BigQuery to Google Nuclear power plans, and everything in between! Welcome to episode 279!  Titles we almost went with this week: AWS SKYNET (Q) now controls the supply chain AWS Supply Chain: Where skynet meets your shopping list Digital Ocean follows Azure with the Premium everything EKS mounts S3  GCP now a nuclear Big query don't hit that iceberg  Big Query Yells: “ICEBERG AHEAD”  The Cloud Pod: Now with 50% more meltdown protection The Cloud Pod radiates excitement over Google's nuclear deal A big thanks to this week's sponsor: We're sponsorless! Want to get your brand, company, or service in front of a very enthusiastic group of cloud news seekers? You've come to the right place! Send us an email or hit us up on our slack channel for more info.  Follow Up 00:46 OpenAI's Newest Possible Threat: Ex-CTO Murati Apologies listeners – paywall article.  Given the recent departure of Ex-CTO Mira Murati from OpenAI, we speculated that she might be starting something new…and the rumors are rumorin'.  Rumors have been running wild since her last day on October 4th, with several people reporting that there has been a lot of churn.  Speculation is that Murati may join former Open AI VP Bret Zoph at his new startup.   It may be easy to steal some people, as the research organization at Open AI is reportedly in upheaval after Liam Fedus’s promotion to lead post-training – several researchers have asked to switch teams.  In addition, Ilya Sutskever, an Open AI co-founder and former chief scientist, also has a new startup.   We'll definitely be keeping an eye on this particular soap opera.  2:00 Jonathan – “I kind wonder what will these other startups bring that’s different than what OpenAI are doing or Anthropic or anybody else. mean, they’re all going to be taking the same training data sets because that’s what’s available. It’s not like they’re going to invent some data from somewhere else and have an edge. I mean, I guess they could do different things like be mindful about licensing.” General News 4:41 Introducing New 48vCPU and 60vCPU Optimized Premium Droplets on DigitalOcean Those raindrops are getting pretty heavy as Digital Ocean announces their new 48vCPU Memory and storage optimized premium droplets, and 60vcpu general purpose and CPU optimized premium droplets.  Droplets are DO's Linux-based virtual machines.   Premium Optimized Droplets are dedicated CPU instances with access to the full hyperthread, as well as 10GBps of outbound data transfer. The 48vCPU boxes have 384GB of memory, and the 60vCPU boxes have 160gb. 6:02 Justin – “I’ve been watchi

From our Sponsors at SimmerGo to TeamSimmer and use the coupon code DEVIATE for 10% on individual course purchases.The Technical Marketing Handbook provides a comprehensive journey through technical marketing principles.A new course is out now! Chrome DevTools for Digital MarketersLatest content from Juliana & SimoArticle: GA4 to Piwik PRO Using Server-side Google Tag Manager by Simo AhavaArticle: Unlocking Real-Time Insights: How does Piwik PRO's Real-Time Dashboarding Feature work? by Juliana JacksonAlso mentioned in the EpisodeKick Point Playbook content consumption tracking recipe from DanaKick Point Playbook Newsletter - The HuddleDana's LinkedIn Learning CoursesGoogle Developers AcademyConnect with Dana DiTomasoDana's LinkedinKick Point Playbook website This podcast is brought to you by Juliana Jackson and Simo Ahava. Intro jingle by Jason Packer and Josh Silverbauer.

Ever wonder how to drive product success when you don't have direct authority over your teams?  In this episode, host Rebecca Kalogeris chats with Leah Zillner, a product manager at Intellum, about the wild ride that is product management. Leah shares her story of transitioning from program management to product, and how Pragmatic Institute's courses helped her navigate the journey. From building market insights through client feedback to using tools like UserPilot, Jira, and BigQuery, Leah has tips that will level up your PM game. She also discusses the internal dynamics of product management, where trust and communication are key (especially when you can't just tell people what to do). Leah talks candidly about learning from mistakes, ditching perfectionism, and building a supportive team culture.  Ready to pick up some insider secrets on how to build relationships, communicate better, and juggle the challenges of product management? This episode has you covered!  For detailed takeaways, show notes, and more, visit: www.pragmaticinstitute.com/resources/podcasts  Pragmatic Institute is the global leader in Product, Data, and Design training and certification programs for working professionals. Learn more at www.pragmaticinstitute.com. 

Simba Khadder is the Founder & CEO of Featureform. He started his ML career in recommender systems where he architected a multi-modal personalization engine that powered 100s of millions of user's experiences. Unpacking 3 Types of Feature Stores // MLOps Podcast #265 with Simba Khadder, Founder & CEO of Featureform. // Abstract Simba dives into how feature stores have evolved and how they now intersect with vector stores, especially in the world of machine learning and LLMs. He breaks down what embeddings are, how they power recommender systems, and why personalization is key to improving LLM prompts. Simba also sheds light on the difference between feature and vector stores, explaining how each plays its part in making ML workflows smoother. Plus, we get into the latest challenges and cool innovations happening in MLOps. // Bio Simba Khadder is the Founder & CEO of Featureform. After leaving Google, Simba founded his first company, TritonML. His startup grew quickly and Simba and his team built ML infrastructure that handled over 100M monthly active users. He instilled his learnings into Featureform's virtual feature store. Featureform turns your existing infrastructure into a Feature Store. He's also an avid surfer, a mixed martial artist, a published astrophysicist for his work on finding Planet 9, and he ran the SF marathon in basketball shoes. // MLOps Jobs board https://mlops.pallet.xyz/jobs // MLOps Swag/Merch https://mlops-community.myshopify.com/ // Related Links Website: featureform.comBigQuery Feature Store // Nicolas Mauti // MLOps Podcast #255: https://www.youtube.com/watch?v=NtDKbGyRHXQ&ab_channel=MLOps.community --------------- ✌️Connect With Us ✌️ ------------- Join our slack community: https://go.mlops.community/slack Follow us on Twitter: @mlopscommunity Sign up for the next meetup: https://go.mlops.community/register Catch all episodes, blogs, newsletters, and more: https://mlops.community/ Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/ Connect with Simba on LinkedIn: https://www.linkedin.com/in/simba-k/ Timestamps: [00:00] Simba's preferred coffee [00:08] Takeaways [02:01] Coining the term 'Embedding' [07:10] Dual Tower Recommender System [10:06] Complexity vs Reliability in AI [12:39] Vector Stores and Feature Stores [17:56] Value of Data Scientists [20:27] Scalability vs Quick Solutions [23:07] MLOps vs LLMOps Debate [24:12] Feature Stores' current landscape [32:02] ML lifecycle challenges and tools [36:16] Feature Stores bundling impact [42:13] Feature Stores and BigQuery [47:42] Virtual vs Literal Feature Store [50:13] Hadoop Community Challenges [52:46] LLM data lifecycle challenges [56:30] Personalization in prompting usage [59:09] Contextualizing company variables [1:03:10] DSPy framework adoption insights [1:05:25] Wrap up

What makes MotherDuck and DuckDB a game-changer for data analytics? Join us as we sit down with Jacob Matson, a renowned expert in SQL Server, dbt, and Excel, who recently became a developer advocate at MotherDuck. During this episode, Jacob shares his compelling journey to MotherDuck, driven by his frequent use of DuckDB for solving data challenges. We explore the unique attributes of DuckDB, comparing it to SQLite for analytics, and uncover its architectural benefits, such as utilizing multi-core machines for parallel query execution. Jacob also sheds light on how MotherDuck is pushing the envelope with their innovative concept of multiplayer analytics.Our discussion takes a deep dive into MotherDuck's innovative tenancy model and how it impacts database workloads, highlighting the use of DuckDB format in Wasm for enhanced data visualization. Jacob explains how this approach offers significant compression and faster query performance, making data visualization more interactive. We also touch on the potential and limitations of replacing traditional BI tools with Mosaic, and where MotherDuck stands in the modern data stack landscape, especially for organizations that don't require the scale of BigQuery or Snowflake. Plus, get a sneak peek into the upcoming Small Data Conference in San Francisco on September 23rd, where we'll explore how small data solutions can address significant problems without relying on big data. Don't miss this episode packed with insights on DuckDB and MotherDuck innovations!Small Data SF Signup  Discount Code: MATSON100What's New In Data is a data thought leadership series hosted by John Kutay who leads data and products at Striim. What's New In Data hosts industry practitioners to discuss latest trends, common patterns for real world data patterns, and analytics success stories.

Google Cloud Data InnovationsThe Big Themes:Integration of unstructured data with AI: Google Cloud is shifting how enterprises leverage their data by integrating unstructured data (which makes up 85-90% of all data) with structured data through its BigQuery multimodal data foundation. This integration allows for a more comprehensive data landscape where AI models can seamlessly access and analyze both types of data. This approach addresses the limitations of traditional data systems and unlocks new potential for AI-driven analytics.The role of partners in maximizing AI and data value: Google Cloud's service partners implement solutions and bring industry best practices to customer environments, while independent software vendors (ISVs) build applications that leverage Google Cloud's data and AI tools. Programs like the Google Cloud Ready (GCR) initiative streamline integrations.Integration challenge: The challenge for organizations lies in connecting disparate data sources, such as operational data from systems like SAP and CRM data from Salesforce, with analytics tools to enable real-time decision-making. Google Cloud addresses this by developing connectors, such as Cortex.The Big Quote: “We are coming to the Third Age in data, which is going to divide data systems. It's not just about having lots of one data type… it's having the broadest possible set of data signals you can bring together. That idea of wide data systems means combining all of your data signals, structured and unstructured, into one unified system."

The newest tech SEO conference is coming to Raleigh, North Carolina, this fall! Guests JR Oakes, Patrick Stox, and Matthew Kay have come together to create an all-new SEO experience, Tech SEO Connect, coming to Raleigh on October 17th & 18th. Don't miss the heavy list of speakers covering core web vitals, Ahrefs Lang, data warehousing, BigQuery, machine learning, and more. In this show, we discuss the origin of Tech SEO Connect with the founders themselves. Learn what makes Tech SEO Connect different from the rest with a diverse content lineup made by technical SEOs for technical SEOs. Get your tickets and mark your calendar as we are all gearing up for the inaugural Tech SEO Connect conference coming this fall. See you there! Key Segments: [00:01:00] Introducing Panelists [00:03:04] The All New TechSEOConnect Conference  [00:07:18] Who is TechSEOConnect Designed For?  [00:12:29] Speakers on the Ballot for Tech SEO Connect [00:13:40] EDGE of the Web Title Sponsor: Site Strategics [00:21:40] Featured Sponsors to Expect at the Conference [00:23:48] What Challenges Arise While Planning an Industry Conference?  [00:24:00] EDGE of The Web Sponsor: Wix   [00:25:47] Unexpected Benefits to Planning Tech SEO Connect  [00:28:06] Tech SEO Connect's Venue Follow Our Guests JR Oakes JR Oakes GitHub Patrick Stox  Matthew Kay  TechSEOConnect Resources:  Tech SEO Connect (Tickets Here)

Nicolas Mauti is an MLOps Engineer from Lyon (France), Working at Malt. BigQuery Feature Store // MLOps Podcast #255 with Nicolas Mauti, Lead MLOps at Malt. // Abstract Need a feature store for your AI/ML applications but overwhelmed by the multitude of options? Think again. In this talk, Nicolas shares how they solved this issue at Malt by leveraging the tools they already had in place. From ingestion to training, Nicolas provides insights on how to transform BigQuery into an effective feature management system. We cover how Nicolas' team designed their feature tables and addressed challenges such as monitoring, alerting, data quality, point-in-time lookups, and backfilling. If you're looking for a simpler way to manage your features without the overhead of additional software, this talk is for you. Discover how BigQuery can handle it all! // Bio Nicolas Mauti is the go-to guy for all things related to MLOps at Malt. With a knack for turning complex problems into streamlined solutions and over a decade of experience in code, data, and ops, he is a driving force in developing and deploying machine learning models that actually work in production. When he's not busy optimizing AI workflows, you can find him sharing his knowledge at the university. Whether it's cracking a tough data challenge or cracking a joke, Nicolas knows how to keep things interesting. // MLOps Jobs board https://mlops.pallet.xyz/jobs // MLOps Swag/Merch https://mlops-community.myshopify.com/ // Related Links Nicolas' Medium - https://medium.com/@nmauti Data Engineering for AI/ML Conference: https://home.mlops.community/home/events/dataengforai --------------- ✌️Connect With Us ✌️ ------------- Join our slack community: https://go.mlops.community/slack Follow us on Twitter: @mlopscommunity Sign up for the next meetup: https://go.mlops.community/register Catch all episodes, blogs, newsletters, and more: https://mlops.community/ Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/ Connect with Nicolas on LinkedIn: https://www.linkedin.com/in/nicolasmauti/?locale=en_US Timestamps: [00:00] Nicolas' preferred beverage [00:35] Takeaways [02:25] Please like, share, leave a review, and subscribe to our MLOps channels! [02:57] BigQuery end goal [05:00] BigQuery pain points [10:14] BigQuery vs Feature Stores [12:54] Freelancing Rate Matching issues [16:43] Post-implementation pain points [19:39] Feature Request Process [20:45] Feature Naming Consistency [23:42] Feature Usage Analysis [26:59] Anomaly detection in data [28:25] Continuous Model Retraining Process [30:26] Model misbehavior detection [33:01] Handling model latency issues [36:28] Accuracy vs The Business [38:59] BigQuery cist-benefit analysis [42:06] Feature stores cost savings [44:09] When not to use BigQuery [46:20] Real-time vs Batch Processing [49:11] Register for the Data Engineering for AI/ML Conference now! [50:14] Wrap up

Data science and engineering workflows often span multiple stages, from warehousing to orchestration, using tools like BigQuery, dbt, and Airbyte. As vision language models (VLMs) advance in multimodal understanding and code generation, VLM-based agents could potentially automate these workflows by generating SQL queries, Python code, and GUI operations. This automation can improve the productivity of experts while democratizing access to large-scale data analysis. In this paper, we introduce Spider2-V, the first multimodal agent benchmark focusing on professional data science and engineering workflows, featuring 494 real-world tasks in authentic computer environments and incorporating 20 enterprise-level professional applications. These tasks, derived from real-world use cases, evaluate the ability of a multimodal agent to perform data-related tasks by writing code and managing the GUI in enterprise data software systems. To balance realistic simulation with evaluation simplicity, we devote significant effort to developing automatic configurations for task setup and carefully crafting evaluation metrics for each task. Furthermore, we supplement multimodal agents with comprehensive documents of these enterprise data software systems. Our empirical evaluation reveals that existing state-of-the-art LLM/VLM-based agents do not reliably automate full data workflows (14.0% success). Even with step-by-step guidance, these agents still underperform in tasks that require fine-grained, knowledge-intensive GUI actions (16.2%) and involve remote cloud-hosted workspaces (10.6%). We hope that Spider2-V paves the way for autonomous multimodal agents to transform the automation of data science and engineering workflow. Our code and data are available at https://spider2-v.github.io. 2024: Ruisheng Cao, Fangyu Lei, Haoyuan Wu, Jixuan Chen, Yeqiao Fu, Hongcheng Gao, Xinzhuang Xiong, Hanchong Zhang, Yuchen Mao, Wenjing Hu, Tianbao Xie, Hongshen Xu, Danyang Zhang, Sida Wang, Ruoxi Sun, Pengcheng Yin, Caiming Xiong, Ansong Ni, Qian Liu, Victor Zhong, Lu Chen, Kai Yu, Tao Yu https://arxiv.org/pdf/2407.10956v1

Highlights from this week's conversation include:David's Background and Career (0:49)Econometrics Work at UPS (3:14)Challenges with Time Series Data and Tools (7:15)Working at Google Cloud (11:28)BigQuery's Significance (13:51)Comparison of Data Warehouse Products (17:23)Learning different cloud platforms (20:17)Coherence in GCP (23:04)Observability and data analysis (32:44)Support for Iceberg format in BigQuery (36:31)AI in Observability (40:25)AI's Role in Observability (43:39)AI and Mental Models (46:04)Final thoughts and takeaways (48:32)The Data Stack Show is a weekly podcast powered by RudderStack, the CDP for developers. Each week we'll talk to data engineers, analysts, and data scientists about their experience around building and maintaining data infrastructure, delivering data and data products, and driving better outcomes across their businesses with data.RudderStack helps businesses make the most out of their customer data while ensuring data privacy and security. To learn more about RudderStack visit rudderstack.com.

From our Sponsors at SimmerGo to TeamSimmer and use the coupon code DEVIATE for 10% on individual course purchases.The Technical Marketing Handbook is live and provides a comprehensive journey through technical marketing principles.A new course is out now! Chrome DevTools for Digital MarketersLatest content from Juliana & SimoArticle: AUTOMATIC PAGE VIEW HITS IN SGTM AFTER CONSENT GRANTED by Simo AhavaArticle: Unlocking Real-Time Insights: How does Piwik PRO's Real-Time Dashboarding Feature work? by Juliana JacksonAlso mentioned in the EpisodeJuliana's NLP Case Study: MediaMonks - AI Customer Voice Analysis Tool for Starbucks EMEAGA4BigqueryConversion Jam 2024Also Asked Tool: AlsoAskedConnect with Jordan PeckLinkedinSnowplow This podcast is brought to you by Juliana Jackson and Simo Ahava. Intro jingle by Jason Packer and Josh Silverbauer.

In this episode of The Marketing Intelligence Show by Supermetrics, Jessica Gondolfo, Head of US Regional Marketing at Supermetrics, is joined by Joe Mineo, the Associate Director of Ads and Analytics at Chatterblast. Together, they'll reveal how to leverage data to navigate the evolving marketing landscape, with a focus on building brand awareness and fostering client trust. Here's what you'll learn: Building brand awareness: Learn how to shift focus from conversions to building brand awareness, especially for younger audiences, with short-form content. Data-driven creatives: Discover how A/B testing unlocks messaging and visuals that resonate with your target audience. Secure data storage: Explore how Supermetrics and BigQuery can help you secure all your historic data (especially as API access changes). Empowering clients: Learn how to build client trust and avoid vendor lock-in with data access through Supermetrics. Tune in to unlock the secrets to data-driven marketing success!

Ibis is a Python library that offers a single data-frame API, from Python, which can run your queries on many different backends. These include databases like Postgres, but also commercial vendors like BigQuery and Snowflake. This ability to control multiple backends from a single API has a lot of use-cases, as well as maintainer challenges, all of which are discussed in this episode. To learn more about Ibis, check out the docs here: https://ibis-project.org/ If you're attending PyCon US this year, you may be interested in Philip's talk: https://us.pycon.org/2024/schedule/presentation/55/ During the podcast, Philip also mentioned a blogpost about DuckDB, here: https://ibis-project.org/posts/why-duckdb/ There was also a dogfooding blogpost, which is this one: https://ibis-project.org/posts/ci-analysis/

We're joined by two BigQuery specialists to cover how to transition from flat-rate to BigQuery Editions and on-demand pricing without overspending.

In today's episode, “Unifying Customer Data for Profits,” we're joined by Rachel Smith, one of the founders of Rex Collective and a seasoned retention marketer to peel back the layers of personalized marketing strategies that are powering profits for e-commerce businesses.With Rachel's we deep dive into understanding why both smaller and larger brands are benefiting from tools like data warehouses, personalization engines, and more. Sean adds to the conversation by stressing the importance of tailored recommendation tags for added personalization and customer engagement.In a rich discussion, we unlock the potential of actioning off a unified customer view, using everything from SMS to paid ads. Rachel invites listeners to explore growth strategies with her team at Rex Collective, while we all ponder the pitfalls and successes in the architecture of modern data systems and the use of tools like Google Analytics and BigQuery.Brands take heed: today's narrative is all about leveraging customer data to not just sell but to build enduring relationships and deliver genuine value. So, tune in and let's translate these numbers into narratives that drive success!Rachel: https://www.linkedin.com/in/rachelgoebel/Rex Collective: https://www.rexcollective.com/

- Paydates and payday schedules in tech firms discussed- The shift in U.S stock trading: Advancement to T+1 instead of T+2- An unexpected discovery: Formula One Mercedes car Lego set- Netflix documentary 'Drive to Survive' recommended for Formula One beginners- Formula One viewed through the lens of an engineer: competition dynamics, changes in rules, strategies of the players- Insight into the upcoming Las Vegas Formula One event- Discussions on AI models: Pros and cons of Llama3, comparing Meta's GPT-3.5 to GPT-4- Key highlights from Google's 2024 Cloud Next Event: AI Agents, AI and the role of BigQuery as a VectorDB- Comparing AI models: ChargerPT vs Llama7b- EV vs gas vehicles: Examining Cybertruck's features, recall event, and travel range- Showcase of Gemini Pro's feature that converts YouTube links into a blog post- Views on owning a Cybertruck: Weighing personal circumstances against the vehicle's features- Discussing EV charging at home: Considering potential cost and utility, possible universal charging standards by Tesla- Job changes revealed: Hosts' anticipation for their new roles at 'Snowflake'- Evaluating 'Hell Divers' game: Discussing PlayStation and Xbox strategies- Episode closure and segue into the next podcast# Links mentioned:- [Join us on our Discord channel](https://discord.gg/T38WpgkHGQ)- [Watch 'Drive to Survive' on Netflix](https://www.netflix.com/title/80204890)

Turning mountains of collected data into actionable steps for effective decision-making in the e-commerce space. From optimizing customer acquisition funnels to dissecting the complexities of multi-touch attribution.In this episode, Jordan West and JJ Reynolds, share valuable insights on actionable information, avoiding analysis paralysis, and the importance of centralizing data for strategic planning. The conversation also takes a personal turn, exploring the hobbies and interests of JJ, including his recent foray into e-bikes and his approach to team building. Listen and learn in this episode!Key takeaways from this episode:Efficiently acquiring new customers involves measuring the marketing efficiency ratio and analyzing customer lifetime value.Effective data management is crucial, including the centralization of data into a warehouse for better reporting and action planning across all departments. Setting clear expectations and comparing them with actual outcomes is important when analyzing data. Mapping out user journeys and aligning them with observed outcomes in the data is key for effective marketing strategies.The importance of taking action on data and iterating, rather than blindly copying what others are doing.Recommended Tools/App:Notion: https://www.notion.so/ ClickUp: https://clickup.com/ Asana: https://asana.com/ Monday: https://monday.com/ Slack: https://slack.com/ Triple Whale: https://www.triplewhale.com/ BigQuery: https://cloud.google.com/  Recommended Podcast/Audiobook:The New One Minute Manager: https://www.kenblanchardbooks.com/book/the-new-one-minute-manager-6/ Today's Guest: JJ Reynolds, Founder of Vision Labs, is a seasoned professional in the world of ecommerce, lead generation, and SaaS. With a keen understanding of the importance of taking action on data, JJ specializes in managing data for clients with 7-9 figure businesses in ecommerce, lead generation, and SaaS. JJ's expertise lies in identifying the best ways to acquire customers, reporting on data effectively, and taking strategic action to drive results.Connect and learn more about JJ and Vision Labs:LinkedIn: https://www.linkedin.com/in/jjreynoldsjr/ Website: https://visionlabs.com/ This episode's sponsor is Finale Inventory- the ultimate solution for accurate and efficient inventory management. Trusted by thousands of brands, Finale offers seamless integrations with over 80 sales channels and platforms. With customizable workflows and reporting features, Finale empowers you to streamline operations and scale your business with ease, preventing overselling and maximizing profitability. Whether you're juggling multiple platforms, expanding your product range or just looking for a way to reduce operational chaos, Finale has the tools you need to succeed. Step into the future of e-commerce with Finale Inventory. Learn more here: Finale Inventory

Summary Maintaining a single source of truth for your data is the biggest challenge in data engineering. Different roles and tasks in the business need their own ways to access and analyze the data in the organization. In order to enable this use case, while maintaining a single point of access, the semantic layer has evolved as a technological solution to the problem. In this episode Artyom Keydunov, creator of Cube, discusses the evolution and applications of the semantic layer as a component of your data platform, and how Cube provides speed and cost optimization for your data consumers. Announcements Hello and welcome to the Data Engineering Podcast, the show about modern data management This episode is brought to you by Datafold – a testing automation platform for data engineers that prevents data quality issues from entering every part of your data workflow, from migration to dbt deployment. Datafold has recently launched data replication testing, providing ongoing validation for source-to-target replication. Leverage Datafold's fast cross-database data diffing and Monitoring to test your replication pipelines automatically and continuously. Validate consistency between source and target at any scale, and receive alerts about any discrepancies. Learn more about Datafold by visiting dataengineeringpodcast.com/datafold (https://www.dataengineeringpodcast.com/datafold). Dagster offers a new approach to building and running data platforms and data pipelines. It is an open-source, cloud-native orchestrator for the whole development lifecycle, with integrated lineage and observability, a declarative programming model, and best-in-class testability. Your team can get up and running in minutes thanks to Dagster Cloud, an enterprise-class hosted solution that offers serverless and hybrid deployments, enhanced security, and on-demand ephemeral test deployments. Go to dataengineeringpodcast.com/dagster (https://www.dataengineeringpodcast.com/dagster) today to get started. Your first 30 days are free! Data lakes are notoriously complex. For data engineers who battle to build and scale high quality data workflows on the data lake, Starburst powers petabyte-scale SQL analytics fast, at a fraction of the cost of traditional methods, so that you can meet all your data needs ranging from AI to data applications to complete analytics. Trusted by teams of all sizes, including Comcast and Doordash, Starburst is a data lake analytics platform that delivers the adaptability and flexibility a lakehouse ecosystem promises. And Starburst does all of this on an open architecture with first-class support for Apache Iceberg, Delta Lake and Hudi, so you always maintain ownership of your data. Want to see Starburst in action? Go to dataengineeringpodcast.com/starburst (https://www.dataengineeringpodcast.com/starburst) and get $500 in credits to try Starburst Galaxy today, the easiest and fastest way to get started using Trino. Your host is Tobias Macey and today I'm interviewing Artyom Keydunov about the role of the semantic layer in your data platform Interview Introduction How did you get involved in the area of data management? Can you start by outlining the technical elements of what it means to have a "semantic layer"? In the past couple of years there was a rapid hype cycle around the "metrics layer" and "headless BI", which has largely faded. Can you give your assessment of the current state of the industry around the adoption/implementation of these concepts? What are the benefits of having a discrete service that offers the business metrics/semantic mappings as opposed to implementing those concepts as part of a more general system? (e.g. dbt, BI, warehouse marts, etc.) At what point does it become necessary/beneficial for a team to adopt such a service? What are the challenges involved in retrofitting a semantic layer into a production data system? evolution of requirements/usage patterns technical complexities/performance and cost optimization What are the most interesting, innovative, or unexpected ways that you have seen Cube used? What are the most interesting, unexpected, or challenging lessons that you have learned while working on Cube? When is Cube/a semantic layer the wrong choice? What do you have planned for the future of Cube? Contact Info LinkedIn (https://www.linkedin.com/in/keydunov/) keydunov (https://github.com/keydunov) on GitHub Parting Question From your perspective, what is the biggest gap in the tooling or technology for data management today? Closing Announcements Thank you for listening! Don't forget to check out our other shows. Podcast.__init__ (https://www.pythonpodcast.com) covers the Python language, its community, and the innovative ways it is being used. The Machine Learning Podcast (https://www.themachinelearningpodcast.com) helps you go from idea to production with machine learning. Visit the site (https://www.dataengineeringpodcast.com) to subscribe to the show, sign up for the mailing list, and read the show notes. If you've learned something or tried out a project from the show then tell us about it! Email hosts@dataengineeringpodcast.com (mailto:hosts@dataengineeringpodcast.com)) with your story. Links Cube (https://cube.dev/) Semantic Layer (https://en.wikipedia.org/wiki/Semantic_layer) Business Objects (https://en.wikipedia.org/wiki/BusinessObjects) Tableau (https://www.tableau.com/) Looker (https://cloud.google.com/looker/?hl=en) Podcast Episode (https://www.dataengineeringpodcast.com/looker-with-daniel-mintz-episode-55/) Mode (https://mode.com/) Thoughtspot (https://www.thoughtspot.com/) LightDash (https://www.lightdash.com/) Podcast Episode (https://www.dataengineeringpodcast.com/lightdash-exploratory-business-intelligence-episode-232/) Embedded Analytics (https://en.wikipedia.org/wiki/Embedded_analytics) Dimensional Modeling (https://en.wikipedia.org/wiki/Dimensional_modeling) Clickhouse (https://clickhouse.com/) Podcast Episode (https://www.dataengineeringpodcast.com/clickhouse-data-warehouse-episode-88/) Druid (https://druid.apache.org/) BigQuery (https://cloud.google.com/bigquery?hl=en) Starburst (https://www.starburst.io/) Pinot (https://pinot.apache.org/) Snowflake (https://www.snowflake.com/en/) Podcast Episode (https://www.dataengineeringpodcast.com/snowflakedb-cloud-data-warehouse-episode-110/) Arrow Datafusion (https://arrow.apache.org/datafusion/) Metabase (https://www.metabase.com/) Podcast Episode (https://www.dataengineeringpodcast.com/metabase-with-sameer-al-sakran-episode-29) Superset (https://superset.apache.org/) Alation (https://www.alation.com/) Collibra (https://www.collibra.com/) Podcast Episode (https://www.dataengineeringpodcast.com/collibra-enterprise-data-governance-episode-188) Atlan (https://atlan.com/) Podcast Episode (https://www.dataengineeringpodcast.com/atlan-data-team-collaboration-episode-179) The intro and outro music is from The Hug (http://freemusicarchive.org/music/The_Freak_Fandango_Orchestra/Love_death_and_a_drunken_monkey/04_-_The_Hug) by The Freak Fandango Orchestra (http://freemusicarchive.org/music/The_Freak_Fandango_Orchestra/) / CC BY-SA (http://creativecommons.org/licenses/by-sa/3.0/)

In this episode, Frank and Steve discuss various news and updates in the cloud industry. They cover topics such as New AMD instances in AzureHolographic stickers

In this conversation, Cinthia and Erica discuss the retirement of Universal Analytics in July 2024 and the transition to GA4. They explore the importance of backing up data and provide options for saving Universal Analytics data, including using tools like BigQuery and third-party solutions. They also discuss baseline reports that should be downloaded and offer recommendations for choosing a data backup solution.   Full episode shownotes for this episode: https://digitalbloomiq.com/seo/ga-universal-data-migration   What You'll Learn in This Episode: Universal Analytics is being retired, and users need to transition to GA4. It is important to back up Universal Analytics data to ensure it can be referenced in the future. Options for saving Universal Analytics data include using tools like BigQuery and third-party solutions. Baseline reports that should be downloaded include traffic, conversions, and revenue data. Resources Mentioned:   BigQuery: https://cloud.google.com/bigquery AnalyticsCanves: https://analyticscanvas.com/ Looker Studio: https://lookerstudio.google.com/   More about Erika Austin:    I am a Digital Marketing Consultant specializing in SEO, Analytics and Advertising. I primarily work with marketers and agencies to improve, prove or scale their marketing efforts to their maximum potential with data-driven insights and emerging technologies. https://erikaaustin.com/   Website Links: Get email updates on all podcast episodes (+ SEO tips, behind the scenes, and early bird offers) : here: https://digitalbloomiq.com/email   90 Day SEO Plan: Your Dream Clients Booking You Overnight! Free webinar training here: https://digitalbloomiq.com/90dayseoplan   More information about the podcast and Digital Bloom IQ:   https://digitalbloomiq.com/podcast https://www.instagram.com/digitalbloomiq/ https://twitter.com/digitalbloomiq https://facebook.com/digitalbloomiq https://www.linkedin.com/in/cinthia-pacheco/ Voice Over, Mixing and Mastering Credits: L. Connor Voice - LConnorvoice@gmail.com Lconnorvoice.com Music Credits:  Music: Kawaii! - Bad Snacks Support by RFM - NCM: https://bit.ly/3f1GFyN

We're writing this one day after the monster release of OpenAI's Sora and Gemini 1.5. We covered this on ‘s ThursdAI space, so head over there for our takes.IRL: We're ONE WEEK away from Latent Space: Final Frontiers, the second edition and anniversary of our first ever Latent Space event! Also: join us on June 25-27 for the biggest AI Engineer conference of the year!Online: All three Discord clubs are thriving. Join us every Wednesday/Friday!Almost 12 years ago, while working at Spotify, Erik Bernhardsson built one of the first open source vector databases, Annoy, based on ANN search. He also built Luigi, one of the predecessors to Airflow, which helps data teams orchestrate and execute data-intensive and long-running jobs. Surprisingly, he didn't start yet another vector database company, but instead in 2021 founded Modal, the “high-performance cloud for developers”. In 2022 they opened doors to developers after their seed round, and in 2023 announced their GA with a $16m Series A.More importantly, they have won fans among both household names like Ramp, Scale AI, Substack, and Cohere, and newer startups like (upcoming guest!) Suno.ai and individual hackers (Modal was the top tool of choice in the Vercel AI Accelerator):We've covered the nuances of GPU workloads, and how we need new developer tooling and runtimes for them (see our episodes with Chris Lattner of Modular and George Hotz of tiny to start). In this episode, we run through the major limitations of the actual infrastructure behind the clouds that run these models, and how Erik envisions the “postmodern data stack”. In his 2021 blog post “Software infrastructure 2.0: a wishlist”, Erik had “Truly serverless” as one of his points:* The word cluster is an anachronism to an end-user in the cloud! I'm already running things in the cloud where there's elastic resources available at any time. Why do I have to think about the underlying pool of resources? Just maintain it for me.* I don't ever want to provision anything in advance of load.* I don't want to pay for idle resources. Just let me pay for whatever resources I'm actually using.* Serverless doesn't mean it's a burstable VM that saves its instance state to disk during periods of idle.Swyx called this Self Provisioning Runtimes back in the day. Modal doesn't put you in YAML hell, preferring to colocate infra provisioning right next to the code that utilizes it, so you can just add GPU (and disk, and retries…):After 3 years, we finally have a big market push for this: running inference on generative models is going to be the killer app for serverless, for a few reasons:* AI models are stateless: even in conversational interfaces, each message generation is a fully-contained request to the LLM. There's no knowledge that is stored in the model itself between messages, which means that tear down / spin up of resources doesn't create any headaches with maintaining state.* Token-based pricing is better aligned with serverless infrastructure than fixed monthly costs of traditional software.* GPU scarcity makes it really expensive to have reserved instances that are available to you 24/7. It's much more convenient to build with a serverless-like infrastructure.In the episode we covered a lot more topics like maximizing GPU utilization, why Oracle Cloud rocks, and how Erik has never owned a TV in his life. Enjoy!Show Notes* Modal* ErikBot* Erik's Blog* Software Infra 2.0 Wishlist* Luigi* Annoy* Hetzner* CoreWeave* Cloudflare FaaS* Poolside AI* Modular Inference EngineChapters* [00:00:00] Introductions* [00:02:00] Erik's OSS work at Spotify: Annoy and Luigi* [00:06:22] Starting Modal* [00:07:54] Vision for a "postmodern data stack"* [00:10:43] Solving container cold start problems* [00:12:57] Designing Modal's Python SDK* [00:15:18] Self-Revisioning Runtime* [00:19:14] Truly Serverless Infrastructure* [00:20:52] Beyond model inference* [00:22:09] Tricks to maximize GPU utilization* [00:26:27] Differences in AI and data science workloads* [00:28:08] Modal vs Replicate vs Modular and lessons from Heroku's "graduation problem"* [00:34:12] Creating Erik's clone "ErikBot"* [00:37:43] Enabling massive parallelism across thousands of GPUs* [00:39:45] The Modal Sandbox for agents* [00:43:51] Thoughts on the AI Inference War* [00:49:18] Erik's best tweets* [00:51:57] Why buying hardware is a waste of money* [00:54:18] Erik's competitive programming backgrounds* [00:59:02] Why does Sweden have the best Counter Strike players?* [00:59:53] Never owning a car or TV* [01:00:21] Advice for infrastructure startupsTranscriptAlessio [00:00:00]: Hey everyone, welcome to the Latent Space podcast. This is Alessio, partner and CTO-in-Residence at Decibel Partners, and I'm joined by my co-host Swyx, founder of Smol AI.Swyx [00:00:14]: Hey, and today we have in the studio Erik Bernhardsson from Modal. Welcome.Erik [00:00:19]: Hi. It's awesome being here.Swyx [00:00:20]: Yeah. Awesome seeing you in person. I've seen you online for a number of years as you were building on Modal and I think you're just making a San Francisco trip just to see people here, right? I've been to like two Modal events in San Francisco here.Erik [00:00:34]: Yeah, that's right. We're based in New York, so I figured sometimes I have to come out to capital of AI and make a presence.Swyx [00:00:40]: What do you think is the pros and cons of building in New York?Erik [00:00:45]: I mean, I never built anything elsewhere. I lived in New York the last 12 years. I love the city. Obviously, there's a lot more stuff going on here and there's a lot more customers and that's why I'm out here. I do feel like for me, where I am in life, I'm a very boring person. I kind of work hard and then I go home and hang out with my kids. I don't have time to go to events and meetups and stuff anyway. In that sense, New York is kind of nice. I walk to work every morning. It's like five minutes away from my apartment. It's very time efficient in that sense. Yeah.Swyx [00:01:10]: Yeah. It's also a good life. So we'll do a brief bio and then we'll talk about anything else that people should know about you. Actually, I was surprised to find out you're from Sweden. You went to college in KTH and your master's was in implementing a scalable music recommender system. Yeah.Erik [00:01:27]: I had no idea. Yeah. So I actually studied physics, but I grew up coding and I did a lot of programming competition and then as I was thinking about graduating, I got in touch with an obscure music streaming startup called Spotify, which was then like 30 people. And for some reason, I convinced them, why don't I just come and write a master's thesis with you and I'll do some cool collaborative filtering, despite not knowing anything about collaborative filtering really. But no one knew anything back then. So I spent six months at Spotify basically building a prototype of a music recommendation system and then turned that into a master's thesis. And then later when I graduated, I joined Spotify full time.Swyx [00:02:00]: So that was the start of your data career. You also wrote a couple of popular open source tooling while you were there. Is that correct?Erik [00:02:09]: No, that's right. I mean, I was at Spotify for seven years, so this is a long stint. And Spotify was a wild place early on and I mean, data space is also a wild place. I mean, it was like Hadoop cluster in the like foosball room on the floor. It was a lot of crude, like very basic infrastructure and I didn't know anything about it. And like I was hired to kind of figure out data stuff. And I started hacking on a recommendation system and then, you know, got sidetracked in a bunch of other stuff. I fixed a bunch of reporting things and set up A-B testing and started doing like business analytics and later got back to music recommendation system. And a lot of the infrastructure didn't really exist. Like there was like Hadoop back then, which is kind of bad and I don't miss it. But I spent a lot of time with that. As a part of that, I ended up building a workflow engine called Luigi, which is like briefly like somewhat like widely ended up being used by a bunch of companies. Sort of like, you know, kind of like Airflow, but like before Airflow. I think it did some things better, some things worse. I also built a vector database called Annoy, which is like for a while, it was actually quite widely used. In 2012, so it was like way before like all this like vector database stuff ended up happening. And funny enough, I was actually obsessed with like vectors back then. Like I was like, this is going to be huge. Like just give it like a few years. I didn't know it was going to take like nine years and then there's going to suddenly be like 20 startups doing vector databases in one year. So it did happen. In that sense, I was right. I'm glad I didn't start a startup in the vector database space. I would have started way too early. But yeah, that was, yeah, it was a fun seven years as part of it. It was a great culture, a great company.Swyx [00:03:32]: Yeah. Just to take a quick tangent on this vector database thing, because we probably won't revisit it but like, has anything architecturally changed in the last nine years?Erik [00:03:41]: I'm actually not following it like super closely. I think, you know, some of the best algorithms are still the same as like hierarchical navigable small world.Swyx [00:03:51]: Yeah. HNSW.Erik [00:03:52]: Exactly. I think now there's like product quantization, there's like some other stuff that I haven't really followed super closely. I mean, obviously, like back then it was like, you know, it's always like very simple. It's like a C++ library with Python bindings and you could mmap big files and into memory and like they had some lookups. I used like this kind of recursive, like hyperspace splitting strategy, which is not that good, but it sort of was good enough at that time. But I think a lot of like HNSW is still like what people generally use. Now of course, like databases are much better in the sense like to support like inserts and updates and stuff like that. I know I never supported that. Yeah, it's sort of exciting to finally see like vector databases becoming a thing.Swyx [00:04:30]: Yeah. Yeah. And then maybe one takeaway on most interesting lesson from Daniel Ek?Erik [00:04:36]: I mean, I think Daniel Ek, you know, he started Spotify very young. Like he was like 25, something like that. And that was like a good lesson. But like he, in a way, like I think he was a very good leader. Like there was never anything like, no scandals or like no, he wasn't very eccentric at all. It was just kind of like very like level headed, like just like ran the company very well, like never made any like obvious mistakes or I think it was like a few bets that maybe like in hindsight were like a little, you know, like took us, you know, too far in one direction or another. But overall, I mean, I think he was a great CEO, like definitely, you know, up there, like generational CEO, at least for like Swedish startups.Swyx [00:05:09]: Yeah, yeah, for sure. Okay, we should probably move to make our way towards Modal. So then you spent six years as CTO of Better. You were an early engineer and then you scaled up to like 300 engineers.Erik [00:05:21]: I joined as a CTO when there was like no tech team. And yeah, that was a wild chapter in my life. Like the company did very well for a while. And then like during the pandemic, yeah, it was kind of a weird story, but yeah, it kind of collapsed.Swyx [00:05:32]: Yeah, laid off people poorly.Erik [00:05:34]: Yeah, yeah. It was like a bunch of stories. Yeah. I mean, the company like grew from like 10 people when I joined at 10,000, now it's back to a thousand. But yeah, they actually went public a few months ago, kind of crazy. They're still around, like, you know, they're still, you know, doing stuff. So yeah, very kind of interesting six years of my life for non-technical reasons, like I managed like three, four hundred, but yeah, like learning a lot of that, like recruiting. I spent all my time recruiting and stuff like that. And so managing at scale, it's like nice, like now in a way, like when I'm building my own startup. It's actually something I like, don't feel nervous about at all. Like I've managed a scale, like I feel like I can do it again. It's like very different things that I'm nervous about as a startup founder. But yeah, I started Modal three years ago after sort of, after leaving Better, I took a little bit of time off during the pandemic and, but yeah, pretty quickly I was like, I got to build something. I just want to, you know. Yeah. And then yeah, Modal took form in my head, took shape.Swyx [00:06:22]: And as far as I understand, and maybe we can sort of trade off questions. So the quick history is started Modal in 2021, got your seed with Sarah from Amplify in 2022. You just announced your Series A with Redpoint. That's right. And that brings us up to mostly today. Yeah. Most people, I think, were expecting you to build for the data space.Erik: But it is the data space.Swyx:: When I think of data space, I come from like, you know, Snowflake, BigQuery, you know, Fivetran, Nearby, that kind of stuff. And what Modal became is more general purpose than that. Yeah.Erik [00:06:53]: Yeah. I don't know. It was like fun. I actually ran into like Edo Liberty, the CEO of Pinecone, like a few weeks ago. And he was like, I was so afraid you were building a vector database. No, I started Modal because, you know, like in a way, like I work with data, like throughout my most of my career, like every different part of the stack, right? Like I thought everything like business analytics to like deep learning, you know, like building, you know, training neural networks, the scale, like everything in between. And so one of the thoughts, like, and one of the observations I had when I started Modal or like why I started was like, I just wanted to make, build better tools for data teams. And like very, like sort of abstract thing, but like, I find that the data stack is, you know, full of like point solutions that don't integrate well. And still, when you look at like data teams today, you know, like every startup ends up building their own internal Kubernetes wrapper or whatever. And you know, all the different data engineers and machine learning engineers end up kind of struggling with the same things. So I started thinking about like, how do I build a new data stack, which is kind of a megalomaniac project, like, because you kind of want to like throw out everything and start over.Swyx [00:07:54]: It's almost a modern data stack.Erik [00:07:55]: Yeah, like a postmodern data stack. And so I started thinking about that. And a lot of it came from like, like more focused on like the human side of like, how do I make data teams more productive? And like, what is the technology tools that they need? And like, you know, drew out a lot of charts of like, how the data stack looks, you know, what are different components. And it shows actually very interesting, like workflow scheduling, because it kind of sits in like a nice sort of, you know, it's like a hub in the graph of like data products. But it was kind of hard to like, kind of do that in a vacuum, and also to monetize it to some extent. I got very interested in like the layers below at some point. And like, at the end of the day, like most people have code to have to run somewhere. So I think about like, okay, well, how do you make that nice? Like how do you make that? And in particular, like the thing I always like thought about, like developer productivity is like, I think the best way to measure developer productivity is like in terms of the feedback loops, like how quickly when you iterate, like when you write code, like how quickly can you get feedback. And at the innermost loop, it's like writing code and then running it. And like, as soon as you start working with the cloud, like it's like takes minutes suddenly, because you have to build a Docker container and push it to the cloud and like run it, you know. So that was like the initial focus for me was like, I just want to solve that problem. Like I want to, you know, build something less, you run things in the cloud and like retain the sort of, you know, the joy of productivity as when you're running things locally. And in particular, I was quite focused on data teams, because I think they had a couple unique needs that wasn't well served by the infrastructure at that time, or like still is in like, in particular, like Kubernetes, I feel like it's like kind of worked okay for back end teams, but not so well for data teams. And very quickly, I got sucked into like a very deep like rabbit hole of like...Swyx [00:09:24]: Not well for data teams because of burstiness. Yeah, for sure.Erik [00:09:26]: So like burstiness is like one thing, right? Like, you know, like you often have this like fan out, you want to like apply some function over very large data sets. Another thing tends to be like hardware requirements, like you need like GPUs and like, I've seen this in many companies, like you go, you know, data scientists go to a platform team and they're like, can we add GPUs to the Kubernetes? And they're like, no, like, that's, you know, complex, and we're not gonna, so like just getting GPU access. And then like, I mean, I also like data code, like frankly, or like machine learning code like tends to be like, super annoying in terms of like environments, like you end up having like a lot of like custom, like containers and like environment conflicts. And like, it's very hard to set up like a unified container that like can serve like a data scientist, because like, there's always like packages that break. And so I think there's a lot of different reasons why the technology wasn't well suited for back end. And I think the attitude at that time is often like, you know, like you had friction between the data team and the platform team, like, well, it works for the back end stuff, you know, why don't you just like, you know, make it work. But like, I actually felt like data teams, you know, or at this point now, like there's so much, so many people working with data, and like they, to some extent, like deserve their own tools and their own tool chains, and like optimizing for that is not something people have done. So that's, that's sort of like very abstract philosophical reason why I started Model. And then, and then I got sucked into this like rabbit hole of like container cold start and, you know, like whatever, Linux, page cache, you know, file system optimizations.Swyx [00:10:43]: Yeah, tell people, I think the first time I met you, I think you told me some numbers, but I don't remember, like, what are the main achievements that you were unhappy with the status quo? And then you built your own container stack?Erik [00:10:52]: Yeah, I mean, like, in particular, it was like, in order to have that loop, right? You want to be able to start, like take code on your laptop, whatever, and like run in the cloud very quickly, and like running in custom containers, and maybe like spin up like 100 containers, 1000, you know, things like that. And so container cold start was the initial like, from like a developer productivity point of view, it was like, really, what I was focusing on is, I want to take code, I want to stick it in container, I want to execute in the cloud, and like, you know, make it feel like fast. And when you look at like, how Docker works, for instance, like Docker, you have this like, fairly convoluted, like very resource inefficient way, they, you know, you build a container, you upload the whole container, and then you download it, and you run it. And Kubernetes is also like, not very fast at like starting containers. So like, I started kind of like, you know, going a layer deeper, like Docker is actually like, you know, there's like a couple of different primitives, but like a lower level primitive is run C, which is like a container runner. And I was like, what if I just take the container runner, like run C, and I point it to like my own root file system, and then I built like my own virtual file system that exposes files over a network instead. And that was like the sort of very crude version of model, it's like now I can actually start containers very quickly, because it turns out like when you start a Docker container, like, first of all, like most Docker images are like several gigabytes, and like 99% of that is never going to be consumed, like there's a bunch of like, you know, like timezone information for like Uzbekistan, like no one's going to read it. And then there's a very high overlap between the files are going to be read, there's going to be like lib torch or whatever, like it's going to be read. So you can also cache it very well. So that was like the first sort of stuff we started working on was like, let's build this like container file system. And you know, coupled with like, you know, just using run C directly. And that actually enabled us to like, get to this point of like, you write code, and then you can launch it in the cloud within like a second or two, like something like that. And you know, there's been many optimizations since then, but that was sort of starting point.Alessio [00:12:33]: Can we talk about the developer experience as well, I think one of the magic things about Modal is at the very basic layers, like a Python function decorator, it's just like stub and whatnot. But then you also have a way to define a full container, what were kind of the design decisions that went into it? Where did you start? How easy did you want it to be? And then maybe how much complexity did you then add on to make sure that every use case fit?Erik [00:12:57]: I mean, Modal, I almost feel like it's like almost like two products kind of glued together. Like there's like the low level like container runtime, like file system, all that stuff like in Rust. And then there's like the Python SDK, right? Like how do you express applications? And I think, I mean, Swix, like I think your blog was like the self-provisioning runtime was like, to me, always like to sort of, for me, like an eye-opening thing. It's like, so I didn't think about like...Swyx [00:13:15]: You wrote your post four months before me. Yeah? The software 2.0, Infra 2.0. Yeah.Erik [00:13:19]: Well, I don't know, like convergence of minds. I guess we were like both thinking. Maybe you put, I think, better words than like, you know, maybe something I was like thinking about for a long time. Yeah.Swyx [00:13:29]: And I can tell you how I was thinking about it on my end, but I want to hear you say it.Erik [00:13:32]: Yeah, yeah, I would love to. So to me, like what I always wanted to build was like, I don't know, like, I don't know if you use like Pulumi. Like Pulumi is like nice, like in the sense, like it's like Pulumi is like you describe infrastructure in code, right? And to me, that was like so nice. Like finally I can like, you know, put a for loop that creates S3 buckets or whatever. And I think like Modal sort of goes one step further in the sense that like, what if you also put the app code inside the infrastructure code and like glue it all together and then like you only have one single place that defines everything and it's all programmable. You don't have any config files. Like Modal has like zero config. There's no config. It's all code. And so that was like the goal that I wanted, like part of that. And then the other part was like, I often find that so much of like my time was spent on like the plumbing between containers. And so my thing was like, well, if I just build this like Python SDK and make it possible to like bridge like different containers, just like a function call, like, and I can say, oh, this function runs in this container and this other function runs in this container and I can just call it just like a normal function, then, you know, I can build these applications that may span a lot of different environments. Maybe they fan out, start other containers, but it's all just like inside Python. You just like have this beautiful kind of nice like DSL almost for like, you know, how to control infrastructure in the cloud. So that was sort of like how we ended up with the Python SDK as it is, which is still evolving all the time, by the way. We keep changing syntax quite a lot because I think it's still somewhat exploratory, but we're starting to converge on something that feels like reasonably good now.Swyx [00:14:54]: Yeah. And along the way you, with this expressiveness, you enabled the ability to, for example, attach a GPU to a function. Totally.Erik [00:15:02]: Yeah. It's like you just like say, you know, on the function decorator, you're like GPU equals, you know, A100 and then or like GPU equals, you know, A10 or T4 or something like that. And then you get that GPU and like, you know, you just run the code and it runs like you don't have to, you know, go through hoops to, you know, start an EC2 instance or whatever.Swyx [00:15:18]: Yeah. So it's all code. Yeah. So one of the reasons I wrote Self-Revisioning Runtimes was I was working at AWS and we had AWS CDK, which is kind of like, you know, the Amazon basics blew me. Yeah, totally. And then, and then like it creates, it compiles the cloud formation. Yeah. And then on the other side, you have to like get all the config stuff and then put it into your application code and make sure that they line up. So then you're writing code to define your infrastructure, then you're writing code to define your application. And I was just like, this is like obvious that it's going to converge, right? Yeah, totally.Erik [00:15:48]: But isn't there like, it might be wrong, but like, was it like SAM or Chalice or one of those? Like, isn't that like an AWS thing that where actually they kind of did that? I feel like there's like one.Swyx [00:15:57]: SAM. Yeah. Still very clunky. It's not, not as elegant as modal.Erik [00:16:03]: I love AWS for like the stuff it's built, you know, like historically in order for me to like, you know, what it enables me to build, but like AWS is always like struggle with developer experience.Swyx [00:16:11]: I mean, they have to not break things.Erik [00:16:15]: Yeah. Yeah. And totally. And they have to build products for a very wide range of use cases. And I think that's hard.Swyx [00:16:21]: Yeah. Yeah. So it's, it's easier to design for. Yeah. So anyway, I was, I was pretty convinced that this, this would happen. I wrote, wrote that thing. And then, you know, I imagine my surprise that you guys had it on your landing page at some point. I think, I think Akshad was just like, just throw that in there.Erik [00:16:34]: Did you trademark it?Swyx [00:16:35]: No, I didn't. But I definitely got sent a few pitch decks with my post on there and it was like really interesting. This is my first time like kind of putting a name to a phenomenon. And I think this is a useful skill for people to just communicate what they're trying to do.Erik [00:16:48]: Yeah. No, I think it's a beautiful concept.Swyx [00:16:50]: Yeah. Yeah. Yeah. But I mean, obviously you implemented it. What became more clear in your explanation today is that actually you're not that tied to Python.Erik [00:16:57]: No. I mean, I, I think that all the like lower level stuff is, you know, just running containers and like scheduling things and, you know, serving container data and stuff. So like one of the benefits of data teams is obviously like they're all like using Python, right? And so that made it a lot easier. I think, you know, if we had focused on other workloads, like, you know, for various reasons, we've like been kind of like half thinking about like CI or like things like that. But like, in a way that's like harder because like you also, then you have to be like, you know, multiple SDKs, whereas, you know, focusing on data teams, you can only, you know, Python like covers like 95% of all teams. That made it a lot easier. But like, I mean, like definitely like in the future, we're going to have others support, like supporting other languages. JavaScript for sure is the obvious next language. But you know, who knows, like, you know, Rust, Go, R, whatever, PHP, Haskell, I don't know.Swyx [00:17:42]: You know, I think for me, I actually am a person who like kind of liked the idea of programming language advancements being improvements in developer experience. But all I saw out of the academic sort of PLT type people is just type level improvements. And I always think like, for me, like one of the core reasons for self-provisioning runtimes and then why I like Modal is like, this is actually a productivity increase, right? Like, it's a language level thing, you know, you managed to stick it on top of an existing language, but it is your own language, a DSL on top of Python. And so language level increase on the order of like automatic memory management. You know, you could sort of make that analogy that like, maybe you lose some level of control, but most of the time you're okay with whatever Modal gives you. And like, that's fine. Yeah.Erik [00:18:26]: Yeah. Yeah. I mean, that's how I look at about it too. Like, you know, you look at developer productivity over the last number of decades, like, you know, it's come in like small increments of like, you know, dynamic typing or like is like one thing because not suddenly like for a lot of use cases, you don't need to care about type systems or better compiler technology or like, you know, the cloud or like, you know, relational databases. And, you know, I think, you know, you look at like that, you know, history, it's a steadily, you know, it's like, you know, you look at the developers have been getting like probably 10X more productive every decade for the last four decades or something that was kind of crazy. Like on an exponential scale, we're talking about 10X or is there a 10,000X like, you know, improvement in developer productivity. What we can build today, you know, is arguably like, you know, a fraction of the cost of what it took to build it in the eighties. Maybe it wasn't even possible in the eighties. So that to me, like, that's like so fascinating. I think it's going to keep going for the next few decades. Yeah.Alessio [00:19:14]: Yeah. Another big thing in the infra 2.0 wishlist was truly serverless infrastructure. The other on your landing page, you called them native cloud functions, something like that. I think the issue I've seen with serverless has always been people really wanted it to be stateful, even though stateless was much easier to do. And I think now with AI, most model inference is like stateless, you know, outside of the context. So that's kind of made it a lot easier to just put a model, like an AI model on model to run. How do you think about how that changes how people think about infrastructure too? Yeah.Erik [00:19:48]: I mean, I think model is definitely going in the direction of like doing more stateful things and working with data and like high IO use cases. I do think one like massive serendipitous thing that happened like halfway, you know, a year and a half into like the, you know, building model was like Gen AI started exploding and the IO pattern of Gen AI is like fits the serverless model like so well, because it's like, you know, you send this tiny piece of information, like a prompt, right, or something like that. And then like you have this GPU that does like trillions of flops, and then it sends back like a tiny piece of information, right. And that turns out to be something like, you know, if you can get serverless working with GPU, that just like works really well, right. So I think from that point of view, like serverless always to me felt like a little bit of like a solution looking for a problem. I don't actually like don't think like backend is like the problem that needs to serve it or like not as much. But I look at data and in particular, like things like Gen AI, like model inference, like it's like clearly a good fit. So I think that is, you know, to a large extent explains like why we saw, you know, the initial sort of like killer app for model being model inference, which actually wasn't like necessarily what we're focused on. But that's where we've seen like by far the most usage. Yeah.Swyx [00:20:52]: And this was before you started offering like fine tuning of language models, it was mostly stable diffusion. Yeah.Erik [00:20:59]: Yeah. I mean, like model, like I always built it to be a very general purpose compute platform, like something where you can run everything. And I used to call model like a better Kubernetes for data team for a long time. What we realized was like, yeah, that's like, you know, a year and a half in, like we barely had any users or any revenue. And like we were like, well, maybe we should look at like some use case, trying to think of use case. And that was around the same time stable diffusion came out. And the beauty of model is like you can run almost anything on model, right? Like model inference turned out to be like the place where we found initially, well, like clearly this has like 10x like better agronomics than anything else. But we're also like, you know, going back to my original vision, like we're thinking a lot about, you know, now, okay, now we do inference really well. Like what about training? What about fine tuning? What about, you know, end-to-end lifecycle deployment? What about data pre-processing? What about, you know, I don't know, real-time streaming? What about, you know, large data munging, like there's just data observability. I think there's so many things, like kind of going back to what I said about like redefining the data stack, like starting with the foundation of compute. Like one of the exciting things about model is like we've sort of, you know, we've been working on that for three years and it's maturing, but like this is so many things you can do like with just like a better compute primitive and also go up to stack and like do all this other stuff on top of it.Alessio [00:22:09]: How do you think about or rather like I would love to learn more about the underlying infrastructure and like how you make that happen because with fine tuning and training, it's a static memory. Like you exactly know what you're going to load in memory one and it's kind of like a set amount of compute versus inference, just like data is like very bursty. How do you make batches work with a serverless developer experience? You know, like what are like some fun technical challenge you solve to make sure you get max utilization on these GPUs? What we hear from people is like, we have GPUs, but we can really only get like, you know, 30, 40, 50% maybe utilization. What's some of the fun stuff you're working on to get a higher number there?Erik [00:22:48]: Yeah, I think on the inference side, like that's where we like, you know, like from a cost perspective, like utilization perspective, we've seen, you know, like very good numbers and in particular, like it's our ability to start containers and stop containers very quickly. And that means that we can auto scale extremely fast and scale down very quickly, which means like we can always adjust the sort of capacity, the number of GPUs running to the exact traffic volume. And so in many cases, like that actually leads to a sort of interesting thing where like we obviously run our things on like the public cloud, like AWS GCP, we run on Oracle, but in many cases, like users who do inference on those platforms or those clouds, even though we charge a slightly higher price per GPU hour, a lot of users like moving their large scale inference use cases to model, they end up saving a lot of money because we only charge for like with the time the GPU is actually running. And that's a hard problem, right? Like, you know, if you have to constantly adjust the number of machines, if you have to start containers, stop containers, like that's a very hard problem. Starting containers quickly is a very difficult thing. I mentioned we had to build our own file system for this. We also, you know, built our own container scheduler for that. We've implemented recently CPU memory checkpointing so we can take running containers and snapshot the entire CPU, like including registers and everything, and restore it from that point, which means we can restore it from an initialized state. We're looking at GPU checkpointing next, it's like a very interesting thing. So I think with inference stuff, that's where serverless really shines because you can drive, you know, you can push the frontier of latency versus utilization quite substantially, you know, which either ends up being a latency advantage or a cost advantage or both, right? On training, it's probably arguably like less of an advantage doing serverless, frankly, because you know, you can just like spin up a bunch of machines and try to satisfy, like, you know, train as much as you can on each machine. For that area, like we've seen, like, you know, arguably like less usage, like for modal, but there are always like some interesting use case. Like we do have a couple of customers, like RAM, for instance, like they do fine tuning with modal and they basically like one of the patterns they have is like very bursty type fine tuning where they fine tune 100 models in parallel. And that's like a separate thing that modal does really well, right? Like you can, we can start up 100 containers very quickly, run a fine tuning training job on each one of them for that only runs for, I don't know, 10, 20 minutes. And then, you know, you can do hyper parameter tuning in that sense, like just pick the best model and things like that. So there are like interesting training. I think when you get to like training, like very large foundational models, that's a use case we don't support super well, because that's very high IO, you know, you need to have like infinite band and all these things. And those are things we haven't supported yet and might take a while to get to that. So that's like probably like an area where like we're relatively weak in. Yeah.Alessio [00:25:12]: Have you cared at all about lower level model optimization? There's other cloud providers that do custom kernels to get better performance or are you just given that you're not just an AI compute company? Yeah.Erik [00:25:24]: I mean, I think like we want to support like a generic, like general workloads in a sense that like we want users to give us a container essentially or a code or code. And then we want to run that. So I think, you know, we benefit from those things in the sense that like we can tell our users, you know, to use those things. But I don't know if we want to like poke into users containers and like do those things automatically. That's sort of, I think a little bit tricky from the outside to do, because we want to be able to take like arbitrary code and execute it. But certainly like, you know, we can tell our users to like use those things. Yeah.Swyx [00:25:53]: I may have betrayed my own biases because I don't really think about modal as for data teams anymore. I think you started, I think you're much more for AI engineers. My favorite anecdotes, which I think, you know, but I don't know if you directly experienced it. I went to the Vercel AI Accelerator, which you supported. And in the Vercel AI Accelerator, a bunch of startups gave like free credits and like signups and talks and all that stuff. The only ones that stuck are the ones that actually appealed to engineers. And the top usage, the top tool used by far was modal.Erik [00:26:24]: That's awesome.Swyx [00:26:25]: For people building with AI apps. Yeah.Erik [00:26:27]: I mean, it might be also like a terminology question, like the AI versus data, right? Like I've, you know, maybe I'm just like old and jaded, but like, I've seen so many like different titles, like for a while it was like, you know, I was a data scientist and a machine learning engineer and then, you know, there was like analytics engineers and there was like an AI engineer, you know? So like, to me, it's like, I just like in my head, that's to me just like, just data, like, or like engineer, you know, like I don't really, so that's why I've been like, you know, just calling it data teams. But like, of course, like, you know, AI is like, you know, like such a massive fraction of our like workloads.Swyx [00:26:59]: It's a different Venn diagram of things you do, right? So the stuff that you're talking about where you need like infinite bands for like highly parallel training, that's not, that's more of the ML engineer, that's more of the research scientist and less of the AI engineer, which is more sort of trying to put, work at the application.Erik [00:27:16]: Yeah. I mean, to be fair to it, like we have a lot of users that are like doing stuff that I don't think fits neatly into like AI. Like we have a lot of people using like modal for web scraping, like it's kind of nice. You can just like, you know, fire up like a hundred or a thousand containers running Chromium and just like render a bunch of webpages and it takes, you know, whatever. Or like, you know, protein folding is that, I mean, maybe that's, I don't know, like, but like, you know, we have a bunch of users doing that or, or like, you know, in terms of, in the realm of biotech, like sequence alignment, like people using, or like a couple of people using like modal to run like large, like mixed integer programming problems, like, you know, using Gurobi or like things like that. So video processing is another thing that keeps coming up, like, you know, let's say you have like petabytes of video and you want to just like transcode it, like, or you can fire up a lot of containers and just run FFmpeg or like, so there are those things too. Like, I mean, like that being said, like AI is by far our biggest use case, but you know, like, again, like modal is kind of general purpose in that sense.Swyx [00:28:08]: Yeah. Well, maybe I'll stick to the stable diffusion thing and then we'll move on to the other use cases for AI that you want to highlight. The other big player in my mind is replicate. Yeah. In this, in this era, they're much more, I guess, custom built for that purpose, whereas you're more general purpose. How do you position yourself with them? Are they just for like different audiences or are you just heads on competing?Erik [00:28:29]: I think there's like a tiny sliver of the Venn diagram where we're competitive. And then like 99% of the area we're not competitive. I mean, I think for people who, if you look at like front-end engineers, I think that's where like really they found good fit is like, you know, people who built some cool web app and they want some sort of AI capability and they just, you know, an off the shelf model is like perfect for them. That's like, I like use replicate. That's great. I think where we shine is like custom models or custom workflows, you know, running things at very large scale. We need to care about utilization, care about costs. You know, we have much lower prices because we spend a lot more time optimizing our infrastructure, you know, and that's where we're competitive, right? Like, you know, and you look at some of the use cases, like Suno is a big user, like they're running like large scale, like AI. Oh, we're talking with Mikey.Swyx [00:29:12]: Oh, that's great. Cool.Erik [00:29:14]: In a month. Yeah. So, I mean, they're, they're using model for like production infrastructure. Like they have their own like custom model, like custom code and custom weights, you know, for AI generated music, Suno.AI, you know, that, that, those are the types of use cases that we like, you know, things that are like very custom or like, it's like, you know, and those are the things like it's very hard to run and replicate, right? And that's fine. Like I think they, they focus on a very different part of the stack in that sense.Swyx [00:29:35]: And then the other company pattern that I pattern match you to is Modular. I don't know.Erik [00:29:40]: Because of the names?Swyx [00:29:41]: No, no. Wow. No, but yeah, yes, the name is very similar. I think there's something that might be insightful there from a linguistics point of view. Oh no, they have Mojo, the sort of Python SDK. And they have the Modular Inference Engine, which is their sort of their cloud stack, their sort of compute inference stack. I don't know if anyone's made that comparison to you before, but like I see you evolving a little bit in parallel there.Erik [00:30:01]: No, I mean, maybe. Yeah. Like it's not a company I'm like super like familiar, like, I mean, I know the basics, but like, I guess they're similar in the sense like they want to like do a lot of, you know, they have sort of big picture vision.Swyx [00:30:12]: Yes. They also want to build very general purpose. Yeah. So they're marketing themselves as like, if you want to do off the shelf stuff, go out, go somewhere else. If you want to do custom stuff, we're the best place to do it. Yeah. Yeah. There is some overlap there. There's not overlap in the sense that you are a closed source platform. People have to host their code on you. That's true. Whereas for them, they're very insistent on not running their own cloud service. They're a box software. Yeah. They're licensed software.Erik [00:30:37]: I'm sure their VCs at some point going to force them to reconsider. No, no.Swyx [00:30:40]: Chris is very, very insistent and very convincing. So anyway, I would just make that comparison, let people make the links if they want to. But it's an interesting way to see the cloud market develop from my point of view, because I came up in this field thinking cloud is one thing, and I think your vision is like something slightly different, and I see the different takes on it.Erik [00:31:00]: Yeah. And like one thing I've, you know, like I've written a bit about it in my blog too, it's like I think of us as like a second layer of cloud provider in the sense that like I think Snowflake is like kind of a good analogy. Like Snowflake, you know, is infrastructure as a service, right? But they actually run on the like major clouds, right? And I mean, like you can like analyze this very deeply, but like one of the things I always thought about is like, why does Snowflake arbitrarily like win over Redshift? And I think Snowflake, you know, to me, one, because like, I mean, in the end, like AWS makes all the money anyway, like and like Snowflake just had the ability to like focus on like developer experience or like, you know, user experience. And to me, like really proved that you can build a cloud provider, a layer up from, you know, the traditional like public clouds. And in that layer, that's also where I would put Modal, it's like, you know, we're building a cloud provider, like we're, you know, we're like a multi-tenant environment that runs the user code. But we're also building on top of the public cloud. So I think there's a lot of room in that space, I think is very sort of interesting direction.Alessio [00:31:55]: How do you think of that compared to the traditional past history, like, you know, you had AWS, then you had Heroku, then you had Render, Railway.Erik [00:32:04]: Yeah, I mean, I think those are all like great. I think the problem that they all faced was like the graduation problem, right? Like, you know, Heroku or like, I mean, like also like Heroku, there's like a counterfactual future of like, what would have happened if Salesforce didn't buy them, right? Like, that's a sort of separate thing. But like, I think what Heroku, I think always struggled with was like, eventually companies would get big enough that you couldn't really justify running in Heroku. So they would just go and like move it to, you know, whatever AWS or, you know, in particular. And you know, that's something that keeps me up at night too, like, what does that graduation risk like look like for modal? I always think like the only way to build a successful infrastructure company in the long run in the cloud today is you have to appeal to the entire spectrum, right? Or at least like the enterprise, like you have to capture the enterprise market. But the truly good companies capture the whole spectrum, right? Like I think of companies like, I don't like Datadog or Mongo or something that were like, they both captured like the hobbyists and acquire them, but also like, you know, have very large enterprise customers. I think that arguably was like where I, in my opinion, like Heroku struggle was like, how do you maintain the customers as they get more and more advanced? I don't know what the solution is, but I think there's, you know, that's something I would have thought deeply if I was at Heroku at that time.Alessio [00:33:14]: What's the AI graduation problem? Is it, I need to fine tune the model, I need better economics, any insights from customer discussions?Erik [00:33:22]: Yeah, I mean, better economics, certainly. But although like, I would say like, even for people who like, you know, needs like thousands of GPUs, just because we can drive utilization so much better, like we, there's actually like a cost advantage of staying on modal. But yeah, I mean, certainly like, you know, and like the fact that VCs like love, you know, throwing money at least used to, you know, add companies who need it to buy GPUs. I think that didn't help the problem. And in training, I think, you know, there's less software differentiation. So in training, I think there's certainly like better economics of like buying big clusters. But I mean, my hope it's going to change, right? Like I think, you know, we're still pretty early in the cycle of like building AI infrastructure. And I think a lot of these companies over in the long run, like, you know, they're, except it may be super big ones, like, you know, on Facebook and Google, they're always going to build their own ones. But like everyone else, like some extent, you know, I think they're better off like buying platforms. And, you know, someone's going to have to build those platforms.Swyx [00:34:12]: Yeah. Cool. Let's move on to language models and just specifically that workload just to flesh it out a little bit. You already said that RAMP is like fine tuning 100 models at once simultaneously on modal. Closer to home, my favorite example is ErikBot. Maybe you want to tell that story.Erik [00:34:30]: Yeah. I mean, it was a prototype thing we built for fun, but it's pretty cool. Like we basically built this thing that hooks up to Slack. It like downloads all the Slack history and, you know, fine-tunes a model based on a person. And then you can chat with that. And so you can like, you know, clone yourself and like talk to yourself on Slack. I mean, it's like nice like demo and it's just like, I think like it's like fully contained modal. Like there's a modal app that does everything, right? Like it downloads Slack, you know, integrates with the Slack API, like downloads the stuff, the data, like just runs the fine-tuning and then like creates like dynamically an inference endpoint. And it's all like self-contained and like, you know, a few hundred lines of code. So I think it's sort of a good kind of use case for, or like it kind of demonstrates a lot of the capabilities of modal.Alessio [00:35:08]: Yeah. On a more personal side, how close did you feel ErikBot was to you?Erik [00:35:13]: It definitely captured the like the language. Yeah. I mean, I don't know, like the content, I always feel this way about like AI and it's gotten better. Like when you look at like AI output of text, like, and it's like, when you glance at it, it's like, yeah, this seems really smart, you know, but then you actually like look a little bit deeper. It's like, what does this mean?Swyx [00:35:32]: What does this person say?Erik [00:35:33]: It's like kind of vacuous, right? And that's like kind of what I felt like, you know, talking to like my clone version, like it's like says like things like the grammar is correct. Like some of the sentences make a lot of sense, but like, what are you trying to say? Like there's no content here. I don't know. I mean, it's like, I got that feeling also with chat TBT in the like early versions right now it's like better, but.Alessio [00:35:51]: That's funny. So I built this thing called small podcaster to automate a lot of our back office work, so to speak. And it's great at transcript. It's great at doing chapters. And then I was like, okay, how about you come up with a short summary? And it's like, it sounds good, but it's like, it's not even the same ballpark as like, yeah, end up writing. Right. And it's hard to see how it's going to get there.Swyx [00:36:11]: Oh, I have ideas.Erik [00:36:13]: I'm certain it's going to get there, but like, I agree with you. Right. And like, I have the same thing. I don't know if you've read like AI generated books. Like they just like kind of seem funny, right? Like there's off, right? But like you glance at it and it's like, oh, it's kind of cool. Like looks correct, but then it's like very weird when you actually read them.Swyx [00:36:30]: Yeah. Well, so for what it's worth, I think anyone can join the modal slack. Is it open to the public? Yeah, totally.Erik [00:36:35]: If you go to modal.com, there's a button in the footer.Swyx [00:36:38]: Yeah. And then you can talk to Erik Bot. And then sometimes I really like picking Erik Bot and then you answer afterwards, but then you're like, yeah, mostly correct or whatever. Any other broader lessons, you know, just broadening out from like the single use case of fine tuning, like what are you seeing people do with fine tuning or just language models on modal in general? Yeah.Erik [00:36:59]: I mean, I think language models is interesting because so many people get started with APIs and that's just, you know, they're just dominating a space in particular opening AI, right? And that's not necessarily like a place where we aim to compete. I mean, maybe at some point, but like, it's just not like a core focus for us. And I think sort of separately, it's sort of a question of like, there's economics in that long term. But like, so we tend to focus on more like the areas like around it, right? Like fine tuning, like another use case we have is a bunch of people, Ramp included, is doing batch embeddings on modal. So let's say, you know, you have like a, actually we're like writing a blog post, like we take all of Wikipedia and like parallelize embeddings in 15 minutes and produce vectors for each article. So those types of use cases, I think modal suits really well for. I think also a lot of like custom inference, like yeah, I love that.Swyx [00:37:43]: Yeah. I think you should give people an idea of the order of magnitude of parallelism, because I think people don't understand how parallel. So like, I think your classic hello world with modal is like some kind of Fibonacci function, right? Yeah, we have a bunch of different ones. Some recursive function. Yeah.Erik [00:37:59]: Yeah. I mean, like, yeah, I mean, it's like pretty easy in modal, like fan out to like, you know, at least like 100 GPUs, like in a few seconds. And you know, if you give it like a couple of minutes, like we can, you know, you can fan out to like thousands of GPUs. Like we run it relatively large scale. And yeah, we've run, you know, many thousands of GPUs at certain points when we needed, you know, big backfills or some customers had very large compute needs.Swyx [00:38:21]: Yeah. Yeah. And I mean, that's super useful for a number of things. So one of my early interactions with modal as well was with a small developer, which is my sort of coding agent. The reason I chose modal was a number of things. One, I just wanted to try it out. I just had an excuse to try it. Akshay offered to onboard me personally. But the most interesting thing was that you could have that sort of local development experience as it was running on my laptop, but then it would seamlessly translate to a cloud service or like a cloud hosted environment. And then it could fan out with concurrency controls. So I could say like, because like, you know, the number of times I hit the GPT-3 API at the time was going to be subject to the rate limit. But I wanted to fan out without worrying about that kind of stuff. With modal, I can just kind of declare that in my config and that's it. Oh, like a concurrency limit?Erik [00:39:07]: Yeah. Yeah.Swyx [00:39:09]: Yeah. There's a lot of control. And that's why it's like, yeah, this is a pretty good use case for like writing this kind of LLM application code inside of this environment that just understands fan out and rate limiting natively. You don't actually have an exposed queue system, but you have it under the hood, you know, that kind of stuff. Totally.Erik [00:39:28]: It's a self-provisioning cloud.Swyx [00:39:30]: So the last part of modal I wanted to touch on, and obviously feel free, I know you're working on new features, was the sandbox that was introduced last year. And this is something that I think was inspired by Code Interpreter. You can tell me the longer history behind that.Erik [00:39:45]: Yeah. Like we originally built it for the use case, like there was a bunch of customers who looked into code generation applications and then they came to us and asked us, is there a safe way to execute code? And yeah, we spent a lot of time on like container security. We used GeoVisor, for instance, which is a Google product that provides pretty strong isolation of code. So we built a product where you can basically like run arbitrary code inside a container and monitor its output or like get it back in a safe way. I mean, over time it's like evolved into more of like, I think the long-term direction is actually I think more interesting, which is that I think modal as a platform where like I think the core like container infrastructure we offer could actually be like, you know, unbundled from like the client SDK and offer to like other, you know, like we're talking to a couple of like other companies that want to run, you know, through their packages, like run, execute jobs on modal, like kind of programmatically. So that's actually the direction like Sandbox is going. It's like turning into more like a platform for platforms is kind of what I've been thinking about it as.Swyx [00:40:45]: Oh boy. Platform. That's the old Kubernetes line.Erik [00:40:48]: Yeah. Yeah. Yeah. But it's like, you know, like having that ability to like programmatically, you know, create containers and execute them, I think, I think is really cool. And I think it opens up a lot of interesting capabilities that are sort of separate from the like core Python SDK in modal. So I'm really excited about C. It's like one of those features that we kind of released and like, you know, then we kind of look at like what users actually build with it and people are starting to build like kind of crazy things. And then, you know, we double down on some of those things because when we see like, you know, potential new product features and so Sandbox, I think in that sense, it's like kind of in that direction. We found a lot of like interesting use cases in the direction of like platformized container runner.Swyx [00:41:27]: Can you be more specific about what you're double down on after seeing users in action?Erik [00:41:32]: I mean, we're working with like some companies that, I mean, without getting into specifics like that, need the ability to take their users code and then launch containers on modal. And it's not about security necessarily, like they just want to use modal as a back end, right? Like they may already provide like Kubernetes as a back end, Lambda as a back end, and now they want to add modal as a back end, right? And so, you know, they need a way to programmatically define jobs on behalf of their users and execute them. And so, I don't know, that's kind of abstract, but does that make sense? I totally get it.Swyx [00:42:03]: It's sort of one level of recursion to sort of be the Modal for their customers.Erik [00:42:09]: Exactly.Swyx [00:42:10]: Yeah, exactly. And Cloudflare has done this, you know, Kenton Vardar from Cloudflare, who's like the tech lead on this thing, called it sort of functions as a service as a service.Erik [00:42:17]: Yeah, that's exactly right. FaSasS.Swyx [00:42:21]: FaSasS. Yeah, like, I mean, like that, I think any base layer, second layer cloud provider like yourself, compute provider like yourself should provide, you know, it's a mark of maturity and success that people just trust you to do that. They'd rather build on top of you than compete with you. The more interesting thing for me is like, what does it mean to serve a computer like an LLM developer, rather than a human developer, right? Like, that's what a sandbox is to me, that you have to redefine modal to serve a different non-human audience.Erik [00:42:51]: Yeah. Yeah, and I think there's some really interesting people, you know, building very cool things.Swyx [00:42:55]: Yeah. So I don't have an answer, but, you know, I imagine things like, hey, the way you give feedback is different. Maybe you have to like stream errors, log errors differently. I don't really know. Yeah. Obviously, there's like safety considerations. Maybe you have an API to like restrict access to the web. Yeah. I don't think anyone would use it, but it's there if you want it.Erik [00:43:17]: Yeah.Swyx [00:43:18]: Yeah. Any other sort of design considerations? I have no idea.Erik [00:43:21]: With sandboxes?Swyx [00:43:22]: Yeah. Yeah.Erik [00:43:24]: Open-ended question here. Yeah. I mean, no, I think, yeah, the network restrictions, I think, make a lot of sense. Yeah. I mean, I think, you know, long-term, like, I think there's a lot of interesting use cases where like the LLM, in itself, can like decide, I want to install these packages and like run this thing. And like, obviously, for a lot of those use cases, like you want to have some sort of control that it doesn't like install malicious stuff and steal your secrets and things like that. But I think that's what's exciting about the sandbox primitive, is like it lets you do that in a relatively safe way.Alessio [00:43:51]: Do you have any thoughts on the inference wars? A lot of providers are just rushing to the bottom to get the lowest price per million tokens. Some of them, you know, the Sean Randomat, they're just losing money and there's like the physics of it just don't work out for them to make any money on it. How do you think about your pricing and like how much premium you can get and you can kind of command versus using lower prices as kind of like a wedge into getting there, especially once you have model instrumented? What are the tradeoffs and any thoughts on strategies that work?Erik [00:44:23]: I mean, we focus more on like custom models and custom code. And I think in that space, there's like less competition and I think we can have a pricing markup, right? Like, you know, people will always compare our prices to like, you know, the GPU power they can get elsewhere. And so how big can that markup be? Like it never can be, you know, we can never charge like 10x more, but we can certainly charge a premium. And like, you know, for that reason, like we can have pretty good margins. The LLM space is like the opposite, like the switching cost of LLMs is zero. If all you're doing is like straight up, like at least like open source, right? Like if all you're doing is like, you know, using some, you know, inference endpoint that serves an open source model and, you know, some other provider comes along and like offers a lower price, you're just going to switch, right? So I don't know, to me that reminds me a lot of like all this like 15 minute delivery wars or like, you know, like Uber versus Lyft, you know, and like maybe going back even further, like I think a lot about like sort of, you know, flip side of this is like, it's actually a positive side, which is like, I thought a lot about like fiber optics boom of like 98, 99, like the other day, or like, you know, and also like the overinvestment in GPU today. Like, like, yeah, like, you know, I don't know, like in the end, like, I don't think VCs will have the return they expected, like, you know, in these things, but guess who's going to benefit, like, you know, is the consumers, like someone's like reaping the value of this. And that's, I think an amazing flip side is that, you know, we should be very grateful, the fact that like VCs want to subsidize these things, which is, you know, like you go back to fiber optics, like there was an extreme, like overinvestment in fiber optics network in like 98. And no one made money who did that. But consumers, you know, got tremendous benefits of all the fiber optics cables that were led, you know, throughout the country in the decades after. I feel something similar abou

In this episode, Lane talks to Alex DeBrie, author of the DynamoDB book. Today's talk covers various aspects such as DynamoDB's comparison with Amazon S3, its benefits, use cases, constraints, and cost considerations, while also covering other AWS and Google Cloud services. Alex also shares his insights into his journey of writing the book on DynamoDB and touches on topics like access patterns, secondary indexes, and billing modes. Alex also shares his professional experiences, including consulting vs freelancing, thoughts of entrepreneurial aspirations, and gives helpful advice for those that are considering pursuing a similar career.Learn back-end development - https://boot.devListen on your favorite podcast player: https://www.backendbanter.fmAlex's Twitter: https://twitter.com/alexbdebrieAlex's Website: https://www.alexdebrie.com(00:00) - Introduction (01:27) - Who is Alex DeBrie? (02:39) - What is DynamoDB? (04:15) - EC2 instance (05:50) - Amazon S3 (06:25) - DynamoDB is more like S3 (07:40) - Difference between DynamoDB and S3 (08:20) - What do we mean when we say NoSQL (10:08) - BigQuery and BigTable (12:31) - Some of DynamoDB's benefits (13:15) - When to use DynamoDB (15:58) - Constraint of number of connections (18:06) - DynamoDB is a multi-tenant service (19:21) - How does DynamoDB shake up against something like MongoDB (22:22) - DynamoDB is opinionated, but it provides good results consistently (25:54) - You can only do certain things in DynamoDB, but they are guaranteed to be fast (26:42) - Relational Databases - Theory vs Practicality (31:08) - How Alex came to write a book about DynamoDB (32:15) - What happens when SQL runs, depends heavily on the system underneath (33:57) - DynamoDB doesn't have a query planner (36:08) - Access patterns (38:04) - Use case for Secondary Indexes (39:43) - Costs of DynamoDB (40:45) - Billing modes for DynamoDB (45:26) - Provisioning and planning for expenses (48:40) - Super Mario 64 Hack (49:34) - What Was Alex's Last Full Time Job (51:02) - Consulting vs Freelancing (52:23) - Does Alex see himself going back to a Full Time Job? (53:07) - Does Alex have any entrepreneurial urges? (54:01) - What you should think about before jumping into freelance/consulting (56:01) - Authority in the consulting world (57:11) - Where to find Alex

On this episode of The Data Chief, Katie Russell, Data Director at OVO Energy shares OVO's transformative journey to become a sustainable energy leader, emphasizing the shift to Google Cloud Platform and a data mesh strategy. The discussion covers OVO's innovative use of generative AI, measuring success through customer savings, and the ongoing challenge of fostering a data-driven culture.Key Moments: OVO's mission and data support [1:05]Data transformation [7:37]Technology modernization [11:17]Data discoverability and data mesh [17:28]Measuring business contribution [22:03]Generative AI and data privacy [27:56]Data-driven culture and trust [30:51]Key Quotes: "We chose Google Cloud Platform as our underlying data platform with BigQuery then as the data warehouse. The thesis being that they practically invented the technology and so should be good at it." “My job is to represent my team, make sure that we're working on the right things, and then, build  trust with the leadership community that we're doing the right things with data for the business.”"I'm thinking that there might actually be a bit of a full circle on data privacy and sharing. I think with ChatGPT being so easy to use with its really human-centered design and with social media ups and downs over the last few years, I'm wondering if there's going to be a revolution in data privacy and data sharing and personal data." Mentions: SQL buddies programPython programGoogle CloudBigQueryMonte CarloAtlanSnowflakeHightouchGenerative AIChatGPTBio: Katie Russell is the Data Director at OVO Energy, leading teams of Data Scientists, Data Engineers and Analysts who are transforming OVO's data capability. As part of a technology led business, leveraging data using artificial intelligence keeps OVO truly innovative, delivering the best possible service for our customers. Katie joined OVO in October 2017 having spent 5 years at ONZO - an energy analytics startup  - as Head of Data Science. During that time she was chuffed to be awarded Big Data Hero by techUK in June 2016 and helped ONZO win multiple awards for their innovative solutions for utilities. Prior to that Katie worked for another analytics start up in the water industry, got a PhD in Mathematical Physics and holds a BA and MMath from the University of Cambridge. 

Join us on a journey as our special guest, Ritu Java, takes us from her beginnings in India to her experiences in Japan, ultimately transforming her into a data-driven entrepreneur. With a unique perspective on the blend of culture and commerce, Ritu shares insights on how she leveraged her expertise in data and analytics to excel in Amazon PPC strategies. You'll also hear her intriguing tales of running an Etsy store from Japan and overcoming the complexities of helping Amazon sellers worldwide. The conversation doesn't stop there. Discover how AI has become a game-changer in running Amazon PPC campaigns as we discuss our personal experiences combining AI with other data sources to optimize campaigns. Listen as we unveil the advantages of using chat GPT for keyword research and translation over traditional methods like Google Translate. This episode offers a unique perspective on integrating AI into workflows and SOPs, driving efficient and effective results. We also underscore the value of incorporating AI into Amazon PPC strategies for successful product launches and campaign management.   To cap off this enlightening conversation, we tackle the future of Amazon selling and the role AI plays in it. From generating keywords for Amazon searches to creating images for sponsored brand ads, we unravel how chat GPT and mid-journey can elevate your selling game. Don't miss out on our tips for creating effective lifestyle photos and the significance of close-up product images. We also shed light on the evolution of Search Query Performance on Amazon and share our strategies for effectively managing and analyzing data. In episode 515 of the Serious Sellers Podcast, Bradley and Ritu discuss: 00:00 - AI Power for E-commerce Sellers 07:54 - Utilizing AI for Amazon Sellers' Success  09:05 - AI in PPC Strategy With Chat GPT 20:52 - Search Term Modifiers and Word Order  23:04 - Enhancing Amazon Ads With AI 31:24 - Generating Posts Using Canva and Amazon  32:19 - Utilizing Search Group Performance Data 33:47 - Optimizing Data Strategy for Efficient Analysis  41:23 - Convert Snapshot Data to Time Series ► Instagram: instagram.com/serioussellerspodcast ► Free Amazon Seller Chrome Extension: https://h10.me/extension ► Sign Up For Helium 10: https://h10.me/signup  (Use SSP10 To Save 10% For Life) ► Learn How To Sell on Amazon: https://h10.me/ft ► Watch The Podcasts On YouTube: youtube.com/@Helium10/videos Transcript Bradley Sutton: Today we've got a first time guest who I think is probably top five in the world these days as far as actionable Amazon strategies, and she's going to give us an absolutely value-packed episode full of tips on generative AI, PPC and more. How cool is that? Pretty cool, I think. How can you get more buyers to leave you Amazon product reviews? By following up with them in a way that's compliant with Amazon terms of service? Bradley Sutton: You can use Helium 10 Follow-Up in order to automatically send out Amazon's request, a review emails, to any customers you want. Not just that, but you can specify when they get the message and even filter out people that you don't want to get that message, such as people who have asked for refunds or maybe ones that you gave discounts to. For more information, visit h10.me forward slash follow-up. You can sign up for a free account or you can sign up for a platinum plan and get 10% off for life by using the discount code SSP10. Hello everybody and welcome to another episode of the Serious Sellers podcast by Helium 10. I'm your host, Bradley Sutton, and this is the show. That's a completely BS free, unscripted and unrehearsed organic conversation about serious strategies for serious sellers of any level in the e-commerce world. We've got a special guest today Ritu. So, first of all, we're going to get into your backstory about how we can even talk in Japanese, because that's something that's crazy. Were you born in Japan or were you born? Ritu: I was born in India, but I lived in Japan for 17 years. Bradley Sutton: So from what age? Ritu: You want to know how old I am. Bradley Sutton: No, no, no. From what age were you living in Japan? Ritu: Mid-20s. Yeah, so mid-20s. Bradley Sutton: Also was, so you didn't go to school in Japan. Ritu: No, I didn't. I went there as an adult. I was working at a company and I take company 17 years. Bradley Sutton: Yes, that means you had to have gone there when you were a child. Then because you can't be over 25 years old. So I don't know what's going on here. Ritu: That is very cute. Bradley Sutton: I was all the reason. I was asking if you grew up because I wore this shirt today. Do you recognize this character here?  What is this? Ritu: Yes Doraemon. Yes, I grew up with Doraemon when I was a little over there, that's awesome. Bradley Sutton: Yes, I grew up with Doraemon when I was a little over there, that's awesome. I know a little bit about you, but I for some reason had this idea that you actually grew up in Japan and that was why you were so fluent in language. Once you go as an adult, it's a little bit harder, unless you really immerse yourself in the culture. Ritu: I did. I really immersed myself in the culture. I went there just for a year, honestly, and ended up staying 17. It's so crazy how that place had such a big impact on me. It was such a stark contrast to where I grew up, which was India. Bradley Sutton: Whereabouts in India. Ritu: In Delhi, the capital city of chaos that's how I describe it from chaotic to super orderly. You can imagine what a difference, that is A stark difference from the world I knew. I was just drawn to the calm and the orderliness of that place. How things were punctual, everything happened as expected, there were no surprises, everything was planned in so much detail, which I kind of liked. I think where I'm at right now is a nice middle ground, because I think I like the chaos. It has energy. It has a certain type of progressive energy that all of us need, especially as entrepreneurs. We need that energy to be able to kind of keep moving forward. But then I also like the organizational skills that I picked up while I was in Japan, because you need that to have good execution. I think best of both worlds is what I'm trying to be at right now, trying to draw from both my cultures. Bradley Sutton: Then did you go to university in India. Ritu: I did. I'm an engineer. I did my electronics engineering from India. I went back to school much later in life. I went back to school in the US and I did a course in data science, which is why I'm very attracted to PPC and data and data analytics and that sort of stuff. Bradley Sutton: When you graduated with the electrical engineering degree, did you start working in India, or is that when you went to Japan? Ritu: Yeah, I started working right away and I started working in India and I worked for an IT company and it was a pretty long stint there as well, like I was very interested in technology right from the start and it kind of aligned with my life's goals and stuff like that. At the time. I mean, little did I know that I would completely switch at a certain point. When I was in Japan I worked for not only the company that I was in India, I kind of went to their Japan office and I started helping them out. But then later on I switched to a more technical role at a school, at a high school, American school in Japan, and then I had my kid and took a break from work and then I kind of dealt in a little bit of entrepreneurship. I started running my own business. I had an Etsy store. Yes, in Japan, while I was in Japan, I started my Etsy business selling jewelry. It was like kind of one of a kind jewelry and I realized that, gosh, it's not enough just to create a listing and people are not going to flock to that listing. So I had to teach myself a whole lot of stuff like marketing advertising. So I learned Facebook ads, Google Ads, blogging, YouTube, all of that stuff. Bradley Sutton: So Etsy in the United States, or is there an Etsy in Japan? Ritu: No, there's an Etsy in the United States, but I was selling on the US market from Japan. So I was producing my stuff there, but I was shipping it worldwide wherever there were shoppers. But shipping costs are exorbitant. Sending stuff from Japan it's very expensive. Yeah, so mostly was attracted to the data side of things. Yes, I have both left and right brains, because the creative side was just all my creations, the jewelry that I made. But then I needed the data science side of things to kind of round things off and make money out of my business, because everything we do here is based on data and I know he's intended the data company. So is PPC Ninja. We might think that we're in the business of selling goods, but actually we're in the business of leveraging data. So that's why it was so important for me to get that knowledge and make sure that I'm kind of ready to go with my own endeavors. Bradley Sutton: Now. So, Etsy was kind of like your first online marketplace. Now, did you ever end up selling on Amazon or did you go straight into software and consulting etc. Ritu: Yeah, so I've never sold on Amazon, but I've helped businesses sell on Amazon. So it's basically the data side of things. So, I only sold on Etsy. I sold on my own website for a bit, but then I have never sold on Amazon myself. But PPC is where I'm focused on. Bradley Sutton: Okay, cool. Now you talked about having an analytical mind, and that's kind of like what you're known for. When you've spoken at events like Billion Dollar Seller Summit and others is especially in the last couple of years, you're one of the go-to people as far as AI and things like that, now me, I'm a little bit behind. I use even on this podcast, we use AI to generate title options and transcripts and things like that, but I would say I'm not one of those full force ahead like, hey, ai is going to replace hours and hours of work. I haven't really adopted it to that effect. So, the typical Amazon seller what are some things that you don't have to be a seven, eight, nine figure seller but just like any Amazon seller if they have not started utilizing AI to help them in their operations or business? What are? Let's take it to that spectrum first. What are some things that you think that any Amazon seller could benefit by utilizing AI? Ritu: Yeah, there's so much. Actually, the magic happens when you start combining things. So AI by itself may not be the be all and all of things, because it's not going to operate in a silo. You've got to combine it with other pieces of data that you have access to. For example, just this morning I was preparing for a new product launch for one of our clients and I'd got all my data from Helium 10. I was at the stage where I have to come up with some keywords for broad match campaigns. I wanted to make sure that all the right keywords are in there, not just the long tail ones with high search volume, but I wanted to make sure that I'm capturing all the seed combinations of important words that make sense. So what I did was I exported the Helium 10 cerebral analysis and I fed it to chat GPT and asked it to come up with two words and three word combinations of seed keywords that would perfectly describe this product. Now what I'm going to do next with that is basically convert that into broad match modifiers, which basically means you add a plus sign in front of all the seeds and then I'm going to create campaigns with it. So that's something that I do at every launch. I generally don't skip that step. It's an important one for me. So, in addition to all the long tail keywords, I will come up with enough seed words that will run at a slightly lower bid but will be like a discovery campaign for me through the broad match modifier channel. So that's kind of one thing that I do. Ritu: Then, like yesterday, I was doing another one for another client, where we have a list of keywords that we discovered from the search query performance report, which is kind of this new, very valuable piece of data that Amazon is giving us these days. So from there I was able to come up with a structure for sponsored brand headline ads and I didn't have to do the work. I just fed that entire list to chat GPT and said, hey, organize this into groups of very related words and then give me a headline ad which is less than 50 characters, because that's the amount Amazon will give us. And then it did that for me. I also gave it one other important instruction, which is to make sure that one of the keywords or a very close variant of that keyword in the group must be included in the title, and that's basically my way of saying, hey, I want this to be a lower funnel ad, not a generic kind of upper funnel ad, because my sponsored brand ads tend to be more focused on ROAS rather than brand discovery and brand awareness. So those are some of the ways that I'm using it almost on a daily basis. I had switched to chat GPT plus a long time ago. I've been paying for it and it's totally worth it. Bradley Sutton: So there's how much is it for somebody to subscribe to? Ritu: that it's about $20 a month. It's not much at all, yeah, it's just $20. And what it gives you is all the beta features, all the new stuff. So right now you can actually upload files very easily. You can upload any kind of file to almost any kind of file to chat GPT and then ask it to analyze, analyze the file and then you can ask it a bunch of questions. So it's just made life so much easier. And I mean I think sky is the limit with what you can do with AI. It's like I always, always feel like I'm not using it enough, even though I'm using it probably quite a bit more than a lot of people, but I still feel cautioned to use it more. Bradley Sutton: Okay, interesting, interesting. So there's some of the ways that you can use it in PPC. Now I remember you presented something. I've seen you speak, you know, various times, but I don't remember which event, this or what it was. That might have been a billion dollars, but where were you doing? You were doing like translation, using like Helium 10 because, like you were doing research, you weren't translating the English keywords. That's obviously a big mistake that some sellers make. Hey, I've got my Amazon USA listing, let me just translate it. Or let me just translate the keywords. No, you need to do the research in that marketplace. So you switch Helium 10 to Amazon Germany, for example, but if you're not a German speaker, you just see all this Deutsch keywords and you don't really know what it means. Or so they're doing it in Amazon Japan and they don't speak Japanese like you, so they might not know. So what's your? I'm not sure if it was AI or just something in Google you were doing to kind of like make that process a little bit easier. Ritu: Yeah. So what we've done is we have integrated chat GPD right into Google Sheets, and we had to write a little bit of code for that. But once we did that, what's happened is that we have these ready to go sheets where we simply change the prompt and add a bunch of keywords and then it will just translate into whatever language, right? So? And I've noticed that any translation done by chat GPD is way better than Google Translate and I've tested it, especially in Japanese, because I can read it. I know that the quality is much better. Ritu: Just to give you an example chat GPD will use the right combinations of Kanji and Hiragana, whereas Google Translate will not. It just doesn't do a great job. And if I tell chat GPD to give me a translation in all four different scripts, that's, kanji as well as Hiragana, Katakana and the Roma G, it will give all those to me. It's a no-brainer to use chat GPD for that sort of thing rather than Google Translate and then other languages as well. Like we're just onboarding this client that has four markets and we have no speakers of those languages on our team. But with chat GPD, we can simply include that into our SOPs, into our workflows and just use those sheets to kind of get the final product out. So it's really great the combination of Helium 10 and chat GPD workflows. They work really well for us. Bradley Sutton: Okay, cool. Now going back a little bit, just remember you were talking about broad match modifiers. There might be people out there who don't know what that means. Can you explain that a little bit? Ritu: Yeah, yeah. So a broad match modifier is a type of broad match, so when you're setting your add up, it'll still be a broad match. However, by simply adding a plus sign before every part of the keyword which means if it's a two word keyword, then both the parts will have a plus sign in front of them what you're gonna ensure is that the buyer search must include those words in exactly that format in order for that match to happen. So this eliminates any kind of kind of synonyms or related words that Amazon might try to kind of connect to, which you don't think need to be there. So at this point, amazon is even replacing exact matches with weird sort of words that it thinks are similar. So we don't want that, because we've done all of the research to find out which exact version of that keyword is giving us the highest search volume, so we wanna stick to it. Ritu: In order to make that happen, we're actually finding ourselves doing more and more work with broad match modifiers, because all the other match types are being weird anymore. Like exact matches are not behaving like exact matches. Same thing with phrase match and broad match anyway, always was a bit too broad and it was always kind of giving you all kinds of weird matches for sponsored brands, but then it started doing the same thing for sponsored products as well, and that makes it a little challenging. It can be wasteful. So yeah, broad match modifiers is a great way of making sure that your matches are clean and that they don't bring in kind of extraneous, superfluous words that you shouldn't be targeting. Bradley Sutton: Do you use that 100% of the time when you have a broad campaign? Ritu: So you always have if it's a three word phrase. Bradley Sutton: You'll put the plus in between each of the. Ritu: Yes, 100% of the time. We've been doing it for the past two years and we actually future proved ourselves because we knew this was coming. It's kind of like Amazon always follows Google. So we knew this was coming because Google introduced broad match modifiers first. Now they've already sunset it. So I don't know where this is gonna end up for Amazon, because what I've heard and I don't wanna just speculate, but what I've heard people say is that Amazon might be moving toward a future where there aren't any match types. There's only a word, there's only a keyword, and then it figures out how to match it the best way. Now it's plausible, especially in this AI world. It's plausible that that might happen. But in the interim, I'm betting on broad match modifiers and exact match. Of course, can't do much about the fact that Amazon isn't treating exact matches the way they ought to be treated, but that's the best we have right now. Bradley Sutton: So what would the difference be between using broad, doing broad target with modifiers compared to phrase for the same, the same, you know, like coffin shelf, like. So if I do coffin plus shelf in broad or coffin shelf in phrase, what's the difference in the potential? You know showings of that keyword. Ritu: Yeah, no, I think the showings of that keyword might totally depend on the bids and they might also depend on relevancy. So it's very hard to predict which of the three match types are gonna win. You know that's been a struggle. I mean you can't really say if you put coffin, what was it? Again coffin shelf. Bradley Sutton: Yeah, coffin shelf. Ritu: Yeah, if you say coffin shelf broad coffin shelf phrase and say coffin shelf exact, what we would want it to do and what would be logical is that if I had a higher bid for exact match, then you know all the searches should come in match through exact match. But that's not always the case. You know, we've seen so much variability there. It also depends on which campaign, you know, starts out those keywords and then each campaign has its own story, its own history. Because let's say, you combine that keyword with a bunch of other keywords and let's say those other keywords got a majority of the early data points, like it started hitting some other words coffin longtail words Before it hit your coffin shelf word, then what happens is that this word starts getting starved of impressions, the other words start to take dominance and these words that get starved of impression give you the false impression that they're not working, whereas it's just a matter of how things started off, like what were the set of searches on that day, on that very moment that Amazon decided to match? Ritu: And then it's going to just take its cues from whatever little data it has in the beginning, because that's all it has to play off of, and then it just keeps giving more and more and more impressions to the early data points and everything else just gets ignored, you know. So it's like a game Like PPC is a game that you know you've got to be able, you've got to be willing to keep playing, trying different things, different ways, moving things, you know, trying it in a different match type, in a different campaign, restarting, stopping, all of that you know. Bradley Sutton: Okay now you know like, for example, if I just do you know, going to this same example, you know coffin shelf, no modifier and broad. You know, yeah, nowadays you know something crazy can come up with, like, you know, spooky decor.You know, potentially it could even come up not even including the word, but ones that are traditional, would be like, you know, coffin shelves for men, coffin shelves for women, but then also it could be coffin shaped shelf, like it could insert a word, or shelf shape like a coffin. You know, like changing the order, but if I put that modifier in there, does that force it, in your experience, to be only longer tail, like it's coffin shelf has to be in there as a phrase and then it's only putting words at the beginning or the end, or still. It could switch it up a little bit. Ritu: Yeah, it will switch it up. So coffin shelf could be shelf coffin even. As long as the word shelf and the word coffin both exist in the match, it will match. Yeah. Bradley Sutton: Okay, going back to Helium 10, now I was looking at, I did it. I still haven't seen your replay of your presentation you did for Helium 10 Elite a few months back. But I was looking at your slides and there was something that you were talking about magnet and seed keywords and just by looking at the slide I couldn't tell what the strategy was. So can you explain what are you doing? I'm not sure if this has to do with chat, gpt or, but just how are you using magnet in a unique way? Ritu: Yeah, so what I do is basically I start off my keyword research by looking at audiences, like who is the right target audience for a product, right? So that's my first step. Now the audience list will help me figure out what words these people use. So if it's a garlic press and let's say there's five different types of people, there could be just regular straight up chefs, there could be restaurant owners, there could be whatever. So there's like five or six different types of people who might use a garlic press. Ritu: Now I ask ChatGPT to tell me all the words that these audiences or avatars are likely to use when they search on Amazon. So I'm actually starting from a suggestion of a seed keyword. That's my starting point, and then I use those seed keywords that chat GPT generates to go and dump that into magnet. And then I use the expand option the second one, not the first one and that basically gives me all of the keywords and their search volumes, and that's what I need Basically. Ritu: I wanna kind of run it by search volume information to figure out if it is really a word that I should be going after. Now I don't always come up with those words, probably because the search volume is too low, in which case I don't need to worry about it, but I can still use that information as broad match modifiers to just generate some sort of discovery. So like, for example, eco-friendly. I don't know if there's any sort of garlic press that's eco-friendly, but let's say someone in that audience wants an eco-friendly garlic press made out of bamboo or whatever. I will still create broad match modifiers that have those important words in that combination so that I can at least start to do some keyword research through an ad rather than through existing search volume data. Bradley Sutton: Okay, cool, switching gears from keywords now to images. I know you've talked about mid-jurdy Canva. Have you played around at all with the new Amazon one that they made kind of for sponsored brands? And then, if so, what's your results? I've had very different, like some of it are absolutely terrible, but then I know that part of it's because I don't really know how to prompt them. I'm not very good at prompting, but what's your experience with the new Amazon AI image generator for sponsored brand ads? Ritu: Yeah, I mean it's not bad for someone who's really struggling with image creation in general, but it's not really usable for every case right? In some cases, it's gonna be hard to come up with the perfect background for your image. The other trouble I have with it is that the product image is too small on the canvas, and that's not how I like my sponsored brand headline ads Generally. This is a tip actually for our listeners when you create a sponsored brand lifestyle photo, the biggest mistake people make is that they fully capture the lifestyle setting in which that product is being used, but then the product itself is so tiny. That's a big mistake. That shouldn't be the way right. The way to do it is to have the product front and center. It has to be blown up right in the middle and then you could maybe suggest what the background is. You might just use suggestive creatives rather than have it in absolute terms. It's being used in the setting that it's being suggested, so for that reason I generally like to request for zoomed in, highly close up type of images so that we can have better conversion rates. Ritu: And there's a story that I just wanna share here real quick. We had one client with a dog product and the product was being used on a dog that was sitting in the lap of a woman on a sofa, and then there's a living room in the background so you can imagine the size of the product. It's like so small you can't see it right. So then what we said to this client was give us a zoomed in image. So then they zoomed right in, so all we see now is the pop and we see the product. Right. So it completely changed the metrics for that ad and then we started using that particular image for many other of their sponsored brand headline ads, and then the rest is history. Ritu: They really started growing after that. But the point is that close up images are more important than pretty images, right? So pretty images anyone can create pretty images. You wanna make them highly converting images and for that reason I might not use the Amazon's AI generated images right away, unless they become better, unless they can kind of keep the product as the hero it needs to be, front and center. Yeah, I'm trying to figure out any prompt that can help me get to that stage, but I'll keep testing. I'm not sure yet. Bradley Sutton: Yeah, so then what outside of Amazon? Then, like I said, I know you're using like mid journey, which is another one that's not too expensive it isn't like 10 bucks a month or something like that to use mid journey, or yeah. So then what if somebody is like all right, you told us what some basic stuff that people how chat GPT for 20 bucks a month can help Amazon sellers. What is something that Amazon sellers of any level can use mid journey for? That's kind of simple and definitely adds value. Ritu: Yeah, I think mid journey is definitely the leader and if you can learn to use it, there's nothing like it yet. But even straight up, chat GPT is now getting pretty good with images, so you can describe whatever you want and then it is connected to dolly in the back and then it generates those images and gives them back to you right in your chat GPT prompt, right. So if you have the paid version, then you can start testing that as well. Bradley Sutton: Okay, so let's say I've got all right, I've got a pretty nice image. You know, maybe it's a white background image or something of my product. Would the first thing I should do with experimenting with AI and mid-journey and things? Would it be making an infographic? Would it be trying to make a lifestyle? Like I remember in the early days of AI, like you could never put a human being in there because they would have like 17 fingers and just crazy faces and stuff like that. But like what should I do then? What kind of images? Or is it not really don't use it for your main images, but use it for, like, the sponsored brand and sponsor display, things like that? Ritu: Yeah, so okay, I think we need to think of images as layers, just like we think of layers in Photoshop. Right, there's layers like a background layer. So if you want just the ambience, the mood, the background, you generate that layer independent of anything else. That's one way of going about it. And then you layer in your product. You have your kind of no background product. Then you can always place it right in the middle, do those sorts of things. So it would probably be a two or three step process where you think of each layer separately, even the humans. You could bring humans in from a different source. You can get humans from there, you can get your backdrop from somewhere else and then you can get your product from your own product images and put them together. That would probably give you the best results. Ritu: But if you tried to have mid-journey to all of that, you might experience some failures there or some surprises with, like you said, 17 fingers and stuff. Now, mid-journey, the latest versions of it are getting better and better, so it's very human-like and it doesn't appear awkward. The facial expressions aren't awkward anymore, so that's good news, just means that we're going in the right direction. It's only gonna get better from here. So I would think of layering as one concept, and then, of course, where you wanna apply it is another thing infographics. I don't think chat, gp or even mid-journey would be good for infographic other than just generating the background for it, because text it still doesn't do a good job with text. You'll have to use some of your other tools for text. So again, it's layering, combining tools and coming up with the concept. So yeah, those are some of the ways in which you can use images. Ritu: Now posts is another interesting one. A lot of people are using mid-journey for generating posts, and that's a good way of generating lots of posts content, because Amazon doesn't allow you to repeat an image twice. So what you can do is you can have Dali or even Canva. I've used Canva AI, which is different from Canva normal. I can explain the difference, but anyway. So Canva AI can generate based on your description of what kind of backgrounds you want, and then you just slap in your photo your kind of hero image on top of it and there you have your posts. It takes barely any time to create like 20 different posts and most people don't realize this, but posts are free advertising. I would highly recommend generating posts on a regular basis and take advantage of it. Bradley Sutton: I've seen them more in search results lately too. Ritu: Posts. Exactly, it's one of those widgets that comes up. Bradley Sutton: That never happened, like six months ago or something. But, now it's right there on page one, so it's important to do, I agree. Ritu: Yeah. Bradley Sutton: All right. So earlier you talked about search group performance. I love search group performance. My self is just like it's stuff that three, four years ago we would have. I would have bet a million dollars that Amazon would never release this kind of data to the public, and Amazon definitely has come a long way. What are some other ways that you're using search group performance, analyzing the data that Amazon gives? Ritu: Yeah, so search group performance. Like you said, it's unbelievable that Amazon is actually sharing this information out, so it's really up to us to take advantage of it as soon as possible. Almost feel like time is of essence here, because everybody's going to have access Everybody has access to that information. But right now most people are in the state of overwhelm. They're like, oh, I have this great data, but I don't know what to do with it. So most people are stuck at that stage. Ritu: But if you want to take the next step, then I would suggest start downloading those reports right away, because these things also get lost. Amazon discontinues things that you think they're going to be giving us forever and forever. For example, the brand analytics data that used to be I don't know millions of rows has certainly been compressed to just 10,000, and so on. So I mean there's a loss there that cannot be replaced. So I would say, number one start downloading your at least your monthly data at the ASIN level and then stitching all that data together, and by stitching I mean maybe putting it into a data warehouse. We use BigQuery in order to bring data in, and the way to stitch it is by making sure that your reports have some extra columns like the date column has to be there Then you have to make sure that you have the brand name in it and you want to make sure that your market is in this, so that when you stitch all that information together, then you can use a single report like a looker studio to dip into the data warehouse and you can basically use switch filters to switch between your different markets. So if you plan your data strategy well, then you will be able to use it more efficiently than just using it in a throwaway style, which most people do. Ritu: Most people go download a report, they look at it, they stare at it and they're like, ok, whatever Done, and it's thrown away. You don't want that. You want a system. You need an ecosystem for managing your data so that you can look at those from time to time. You get a month over month review. You get a month over month trend. You can see if anything has lost its search volume over time. It's so easy to check that at a search term level. Once you have stitched all that information together and is available in maybe something like a looker studio, how about something that's good? Bradley Sutton: it's important to understand the you know, like how to get started and not just like, all right, let me. Let me just look at search career performance or this data, just, you know, in the UI on on Amazon. But then what's the next step? Now I've got everything in my data warehouse and stuff like, for example, me. One of the things I like to look at in search career performance is comparing the conversion rate by the keyword for for just the overall niche, compared to my own. You know my own conversion rate. But you know, I think that's probably one of the most no brainer things. What are some other maybe not so common things that you're looking at when, when you get all of that data into your, your data warehouse, and start you know, start looking up stuff? Ritu: Yeah. So one of the things that I find really interesting is the average price per search term. So this is you know, amazon gives you the average price and that, basically, is a good indication of whether that search term is going for cheaper products or is it going for slightly more expensive products. Just to give you an example, let's say you have the word lotion right Now. You have a $50 lotion by L'Oreal, maybe, and you have a $5 drugstore brand Same thing, selling lotion. But if you're going after, if you're looking at the search term lotion, whatever, daily lotion or whatever and if you see that the average price for that search term is going at $6, let's say that's the average price of the product being sold. That is telling me that, no matter what I do to compete on that, on that search term, it's going to be hard because I'm going to be competing with lots and lots of cheaper brands. So we actually have filters on our search terms or search query reports, so that we only look at those searches that are in the ballpark of our products price point. That basically eliminates a lot of the noise, because otherwise you might be led into thinking that gosh, this is a great keyword and then you spend lots of money on it and ends up being a high cost scenario. You don't want that. So you look at both of the things one that you mentioned, which is what we call strength, keyword strength, which is determined as a ratio of purchase share and impression share. If you can get that ratio to be above one, then that's a good keyword. That is strong, inherently strong, because you're winning more of the purchase share than you're winning of the market, which basically puts it in a good spot. Ritu: And then the second one would be the filter on price. The third filter I would put is search volume, because, again, we don't want noisy, insignificant terms to distract us. And I think the fourth filter I would put there is data sufficiency, like how many sales have you had for that keyword over that period of time? So yeah, those would be the four filters to kind of get everything else out. And then, yeah, I mean that would be our way of figuring out which search terms are good. Then the other use cases of that would be to stitch that data with your ad data. So when you stitch those two together you can find gaps in a systematic sort of way, not just like a one off, throw away kind of way, where it's always being merged and it's always coming together and you can always see these are the ones that I'm not advertising yet. And then, yeah, I think those were the two main ones. Ritu: The third, slightly more advanced one, is when you want to figure out if a search term is good for product A, product B, product C, product D off your catalog because they might be sharing those keywords. Then you can see relative strength across your different products and see where you want to channel your information. Now that comes with the caveat, and that caveat is that there's a very high halo sales ratio on Amazon, which means you might be directing traffic to one of your product variations and something else is actually getting picked up eventually. So you need to know all of the. You need to know all those pieces in order to make the right decision and essentially in terms of using your, your traffic source as a fire hose, literally, and saying, okay, I want to direct it to this product and not to this product. Unless you know what the halo sales are, you could be off. Bradley Sutton: Yeah. Yeah, well really great stuff. Now, before we get into your last strategy you know, maybe it could be a PPC strategy, since that's your specialty how can people reach out to you if they, you know? How can they find you on the interwebs if they want to? You know, get some help with some of the stuff that you've been talking about today. Ritu: Yeah, absolutely so. I'm on LinkedIn. I'm pretty active there, so just look up my full name, Ritu Java, and you should be able to find me there and just say hi and I'll be happy to help. Yeah, and other ways, you can just reach out to our website, ppcninja.com or anywhere else. You see me. Bradley Sutton: Awesome, awesome. Now we have some of we do on our show. We call it TST. That's the 30 second tip. So you know you've been giving us lots of great tips and strategies, but what's like a hard hitting one you can give us in 30 seconds or 60 seconds or less. I'm not going to cut you off, go ahead. Ritu: So I think that you know we're all sitting on tons and tons of data and we don't know how to use it. I would suggest start thinking of strategies to use your data by connecting them up. Every piece of data that we get from Amazon or other sources, whether it's keyword rank tracking or search volume data, or your ads data or organic data. Also, you know competitor data and stuff like that. It's in different locations, it's hiding behind wall gardens and stuff like that. Ritu: You want to figure out a system to bring it all together, and I would recommend using a data warehousing strategy to start bringing everything together so that you can start looking at it holistically. So I would recommend start to think of simple ways in which you can convert your snapshot data into time series. That that would be my advice, and time series is basically for people who don't understand that. It's basically assigning dates to all your downloads. If you're downloading a business report, make sure you add a column and put the date there so that that becomes a way of identifying when that event happened. When you're connecting so many pieces of data together. Bradley Sutton: Awesome, Awesome Well thank you very much. Thank you so much for your time. Ritu: Than you so much Bradley. Bradley Sutton: This was really awesome, awesome and we'll definitely be having you back on the show sometime next year to get your latest strategies. Ritu: Awesome, we'll look forward to that. Take care, Bradley, have a good one.  

Summary Building a data platform that is enjoyable and accessible for all of its end users is a substantial challenge. One of the core complexities that needs to be addressed is the fractal set of integrations that need to be managed across the individual components. In this episode Tobias Macey shares his thoughts on the challenges that he is facing as he prepares to build the next set of architectural layers for his data platform to enable a larger audience to start accessing the data being managed by his team. Announcements Hello and welcome to the Data Engineering Podcast, the show about modern data management Introducing RudderStack Profiles. RudderStack Profiles takes the SaaS guesswork and SQL grunt work out of building complete customer profiles so you can quickly ship actionable, enriched data to every downstream team. You specify the customer traits, then Profiles runs the joins and computations for you to create complete customer profiles. Get all of the details and try the new product today at dataengineeringpodcast.com/rudderstack (https://www.dataengineeringpodcast.com/rudderstack) You shouldn't have to throw away the database to build with fast-changing data. You should be able to keep the familiarity of SQL and the proven architecture of cloud warehouses, but swap the decades-old batch computation model for an efficient incremental engine to get complex queries that are always up-to-date. With Materialize, you can! It's the only true SQL streaming database built from the ground up to meet the needs of modern data products. Whether it's real-time dashboarding and analytics, personalization and segmentation or automation and alerting, Materialize gives you the ability to work with fresh, correct, and scalable results — all in a familiar SQL interface. Go to dataengineeringpodcast.com/materialize (https://www.dataengineeringpodcast.com/materialize) today to get 2 weeks free! Developing event-driven pipelines is going to be a lot easier - Meet Functions! Memphis functions enable developers and data engineers to build an organizational toolbox of functions to process, transform, and enrich ingested events “on the fly” in a serverless manner using AWS Lambda syntax, without boilerplate, orchestration, error handling, and infrastructure in almost any language, including Go, Python, JS, .NET, Java, SQL, and more. Go to dataengineeringpodcast.com/memphis (https://www.dataengineeringpodcast.com/memphis) today to get started! Data lakes are notoriously complex. For data engineers who battle to build and scale high quality data workflows on the data lake, Starburst powers petabyte-scale SQL analytics fast, at a fraction of the cost of traditional methods, so that you can meet all your data needs ranging from AI to data applications to complete analytics. Trusted by teams of all sizes, including Comcast and Doordash, Starburst is a data lake analytics platform that delivers the adaptability and flexibility a lakehouse ecosystem promises. And Starburst does all of this on an open architecture with first-class support for Apache Iceberg, Delta Lake and Hudi, so you always maintain ownership of your data. Want to see Starburst in action? Go to dataengineeringpodcast.com/starburst (https://www.dataengineeringpodcast.com/starburst) and get $500 in credits to try Starburst Galaxy today, the easiest and fastest way to get started using Trino. Your host is Tobias Macey and today I'll be sharing an update on my own journey of building a data platform, with a particular focus on the challenges of tool integration and maintaining a single source of truth Interview Introduction How did you get involved in the area of data management? data sharing weight of history existing integrations with dbt switching cost for e.g. SQLMesh de facto standard of Airflow Single source of truth permissions management across application layers Database engine Storage layer in a lakehouse Presentation/access layer (BI) Data flows dbt -> table level lineage orchestration engine -> pipeline flows task based vs. asset based Metadata platform as the logical place for horizontal view Contact Info LinkedIn (https://linkedin.com/in/tmacey) Website (https://www.dataengineeringpodcast.com) Parting Question From your perspective, what is the biggest gap in the tooling or technology for data management today? Closing Announcements Thank you for listening! Don't forget to check out our other shows. Podcast.__init__ (https://www.pythonpodcast.com) covers the Python language, its community, and the innovative ways it is being used. The Machine Learning Podcast (https://www.themachinelearningpodcast.com) helps you go from idea to production with machine learning. Visit the site (https://www.dataengineeringpodcast.com) to subscribe to the show, sign up for the mailing list, and read the show notes. If you've learned something or tried out a project from the show then tell us about it! Email hosts@dataengineeringpodcast.com (mailto:hosts@dataengineeringpodcast.com)) with your story. To help other people find the show please leave a review on Apple Podcasts (https://podcasts.apple.com/us/podcast/data-engineering-podcast/id1193040557) and tell your friends and co-workers Links Monologue Episode On Data Platform Design (https://www.dataengineeringpodcast.com/data-platform-design-episode-268) Monologue Episode On Leaky Abstractions (https://www.dataengineeringpodcast.com/abstractions-and-technical-debt-episode-374) Airbyte (https://airbyte.com/) Podcast Episode (https://www.dataengineeringpodcast.com/airbyte-open-source-data-integration-episode-173/) Trino (https://trino.io/) Dagster (https://dagster.io/) dbt (https://www.getdbt.com/) Snowflake (https://www.snowflake.com/en/) BigQuery (https://cloud.google.com/bigquery) OpenMetadata (https://open-metadata.org/) OpenLineage (https://openlineage.io/) Data Platform Shadow IT Episode (https://www.dataengineeringpodcast.com/shadow-it-data-analytics-episode-121) Preset (https://preset.io/) LightDash (https://www.lightdash.com/) Podcast Episode (https://www.dataengineeringpodcast.com/lightdash-exploratory-business-intelligence-episode-232/) SQLMesh (https://sqlmesh.readthedocs.io/) Podcast Episode (https://www.dataengineeringpodcast.com/sqlmesh-open-source-dataops-episode-380) Airflow (https://airflow.apache.org/) Spark (https://spark.apache.org/) Flink (https://flink.apache.org/) Tabular (https://tabular.io/) Iceberg (https://iceberg.apache.org/) Open Policy Agent (https://www.openpolicyagent.org/) The intro and outro music is from The Hug (http://freemusicarchive.org/music/The_Freak_Fandango_Orchestra/Love_death_and_a_drunken_monkey/04_-_The_Hug) by The Freak Fandango Orchestra (http://freemusicarchive.org/music/The_Freak_Fandango_Orchestra/) / CC BY-SA (http://creativecommons.org/licenses/by-sa/3.0/)

John Wynkoop, Cloud Economist & Platypus Herder at The Duckbill Group, joins Corey on Screaming in the Cloud to discuss why he decided to make a career move and become an AWS billing consultant. Corey and John discuss how once you're deeply familiar with one cloud provider, those skills become transferable to other cloud providers as well. John also shares the trends he has seen post-pandemic in the world of cloud, including the increased adoption of a multi-cloud strategy and the need for costs control even for VC-funded start-ups. About JohnWith over 25 years in IT, John's done almost every job in the industry, from running cable and answering helpdesk calls to leading engineering teams and advising the C-suite. Before joining The Duckbill Group, he worked across multiple industries including private sector, higher education, and national defense. Most recently he helped IGNW, an industry leading systems integration partner, get acquired by industry powerhouse CDW. When he's not helping customers spend smarter on their cloud bill, you can find him enjoying time with his family in the beautiful Smoky Mountains near his home in Knoxville, TN.Links Referenced: The Duckbill Group: https://duckbillgroup.com LinkedIn: https://www.linkedin.com/in/jlwynkoop/ TranscriptAnnouncer: Hello, and welcome to Screaming in the Cloud with your host, Chief Cloud Economist at The Duckbill Group, Corey Quinn. This weekly show features conversations with people doing interesting work in the world of cloud, thoughtful commentary on the state of the technical world, and ridiculous titles for which Corey refuses to apologize. This is Screaming in the Cloud.Corey: Welcome to Screaming in the Cloud. I'm Corey Quinn. And the times, they are changing. My guest today is John Wynkoop. John, how are you?John: Hey, Corey, I'm doing great. Thanks for having me.Corey: So, big changes are afoot for you. You've taken a new job recently. What are you doing now?John: Well [laugh], so I'm happy to say I have joined The Duckbill Group as a cloud economist. So, came out of the big company world, and have dived back in—or dove back into the startup world.Corey: It's interesting because when we talk to those big companies, they always identify us as oh, you're a startup, which is hilarious on some level because our AWS account hangs out in AWS's startup group, but if you look at the spend being remarkably level from month to month to month to year to year to year, they almost certainly view us as they're a startup, but they suck at it. They completely failed. And so, many of the email stuff that you get from them presupposes that you're venture-backed, that you're trying to conquer the entire world. We don't do that here. We have this old-timey business model that our forebears would have understood of, we make more money than we spend every month and we continue that trend for a long time. So first, thanks for joining us, both on the show and at the company. We like having you around.John: Well, thanks. And yeah, I guess that's—maybe a startup isn't the right word to describe what we do here at The Duckbill Group, but as you said, it seems to fit into the industry classification. But that was one of the things I actually really liked about the—that was appealing about joining the team was, we do spend less than we make and we're not after hyper-growth and we're not trying to consume everything.Corey: So, it's interesting when you put a job description out into the world and you see who applies—and let's be clear, for those who are unaware, job descriptions are inherently aspirational shopping lists. If you look at a job description and you check every box on the thing and you've done all the things they want, the odds are terrific you're going to be bored out of your mind when you wind up showing up to do these… whatever that job is. You should be learning stuff and growing. At least that's always been my philosophy to it. One of the interesting things about you is that you checked an awful lot of boxes, but there is one that I think would cause people to raise an eyebrow, which is, you're relatively new to the fun world of AWS.John: Yeah. So, obviously I, you know, have been around the block a few times when it comes to cloud. I've used AWS, built some things in AWS, but I wouldn't have classified myself as an AWS guru by any stretch of the imagination. I spent the last probably three years working in Google Cloud, helping customers build and deploy solutions there, but I do at least understand the fundamentals of cloud, and more importantly—at least for our customers—cloud costs because at the end of the day, they're not all that different.Corey: I do want to call out that you have a certain humility to you which I find endearing. But you're not allowed to do that here; I will sing your praises for you. Before they deprecated it like they do almost everything else, you were one of the relatively few Google Cloud Certified Fellows, which was sort of like their Heroes program only, you know, they killed it in favor of something else like there's a Champion program or whatnot. You are very deep in the world of both Kubernetes and Google Cloud.John: Yeah. So, there was a few of us that were invited to come out and help Google pilot that program in, I believe it was 2019, and give feedback to help them build the Cloud Fellows Program. And thankfully, I was selected based on some of our early experience with Anthos, and specifically, it was around Certified Fellow in what they call hybrid multi-cloud, so it was experience around Anthos. Or at the time, they hadn't called it Anthos; they were calling it CSP or Cloud Services Platform because that's not an overloaded acronym. So yeah, definitely, was very humbled to be part of that early on.I think the program, as you said, grew to about 70 or so maybe 100 certified individuals before they transitioned—not killed—transitioned to that program into the Cloud Champions program. So, those folks are all still around, myself included. They've just now changed the moniker. But we all get to use the old title still as well, so that's kind of cool.Corey: I have to ask, what would possess you to go from being one of the best in the world at using Google Cloud over here to our corner of the AWS universe? Because the inverse, if I were to somehow get ejected from here—which would be a neat trick, but I'm sure it's theoretically possible—like, “What am I going to do now?” I would almost certainly wind up doing something in the AWS ecosystem, just due to inertia, if nothing else. You clearly didn't see things quite that way. Why make the switch?John: Well, a couple of different reasons. So, being at a Google partner presents a lot of challenges and one of the things that was supremely interesting about coming to Duckbill is that we're independent. So, we're not an AWS partner. We are an independent company that is beholden only to our customers. And there isn't anything like that in the Google ecosystem today.There's, you know, there's Google partners and then there's Google customers and then there's Google. So, that was part of the appeal. And the other thing was, I enjoy learning new things, and honestly, learning, you know, into the depths of AWS cost hell is interesting. There's a lot to learn there and there's a lot of things that we can extract and use to help customers spend less. So, that to me was super interesting.And then also, I want to help build an organization. So, you know, I think what we're doing here at The Duckbill Group is cool and I think that there's an opportunity to grow our services portfolio, and so I'm excited to work with the leadership team to see what else we can bring to market that's going to help our customers, you know, not just with cost optimization, not just with contract negotiation, but you know, through the lifecycle of their AWS… journey, I guess we'll call it.Corey: It's one of those things where I always have believed, on some level, that once you're deep in a particular cloud provider, if there's reason for it, you can rescale relatively quickly to a different provider. There are nuances—deep nuances—that differ from provider to provider, but the underlying concepts generally all work the same way. There's only so many ways you can have data go from point A to point B. There's only so many ways to spin up a bunch of VMs and whatnot. And you're proof-positive that theory was correct.You'd been here less than a week before I started learning nuances about AWS billing from you. I think it was something to do with the way that late fees are assessed when companies don't pay Amazon as quickly as Amazon desires. So, we're all learning new things constantly and no one stuffs this stuff all into their head. But that, if nothing else, definitely cemented that yeah, we've got the right person in the seat.John: Yeah, well, thanks. And certainly, the deeper you go on a specific cloud provider, things become fresh in your memory, you know, other cached so to speak. So, coming up to speed on AWS has been a little bit more documentation reading than it would have been, if I were, say, jumping right into a GCP engagement. But as he said, at the end of the day, there's a lot of similarities. Obviously understanding the nuances of, for example, account organization versus, you know, GCP's Project and Folders. Well, that's a substantial difference and so there's a lot of learning that has to happen.Thankfully, you know, all these companies, maybe with the exception of Oracle, have done a really good job of documenting all of the concepts in their publicly available documentation. And then obviously, having a team of experts here at The Duckbill Group to ask stupid questions of doesn't hurt. But definitely, it's not as hard to come up to speed as one may think, once you've got it understood in one provider.Corey: I took a look recently and was kind of surprised to discover that I've been doing this—as an independent consultant prior to the formation of The Duckbill Group—for seven years now. And it's weird, but I've gone through multiple industry cycles and changes as a part of this. And it feels like I haven't been doing it all that long, but I guess I have. One thing that's definitely changed is that it used to be that companies would basically pick one provider and almost everything would live there. At any reasonable point of scale, everyone is using multiple things.I see Google in effectively every client that we have. It used to be that going to Google Cloud Next was a great place to hang out with AWS customers. But these days, it's just as true to say that a great reason to go to re:Invent is to hang out with Google Cloud customers. Everyone uses everything, and that has become much more clear over the last few years. What have you seen change over the… I guess, since the start of the pandemic, just in terms of broad cycles?John: Yeah. So, I think there's a couple of different trends that we're seeing. Obviously, one is that as you said, especially as large enterprises make moves to the cloud, you see independent teams or divisions within a given organization leveraging… maybe not the right tool for the job because I think that there's a case to be made for swapping out a specific set of tools and having your team learn it, but we do see what I like to refer to as tool fetishism where you get a team that's super, super deep into BigQuery and they're not interested in moving to Redshift, or Snowflake, or a competitor. So, you see, those start to crop up within large organizations where the distributed—the purchasing power, rather—is distributed. So, that's one of the trends is the multi-cloud adoption.And I think the big trend that I like to emphasize around multi-cloud is, just because you can run it anywhere doesn't mean you should run it everywhere. So Kubernetes, as you know, right, as it took off 2019 timeframe, 2020, we started to see a lot of people using that as an excuse to try to run their production application in two, three public cloud providers and on-prem. And unless you're a SaaS customer—or SaaS company with customers in every cloud, there's very little reason to do that. But having that flexibility—that's the other one, is we've seen that AWS has gotten a little difficult to negotiate with, or maybe Google and Microsoft have gotten a little bit more aggressive. So obviously, having that flexibility and being able to move your workloads, that was another big trend.Corey: I'm seeing a change in things that I had taken as givens, back when I started. And that's part of the reason, incidentally, I write the Last Week in AWS newsletter because once you learn a thing, it is very easy not to keep current with that thing, and things that are not possible today will be possible tomorrow. How do you keep abreast of all of those changes? And the answer is to write a deeply sarcastic newsletter that gathers in everything from the world of AWS. But I don't recommend that for most people. One thing that I've seen in more prosaic terms that you have a bit of background in is that HPC on cloud was, five, six years ago, met with, “Oh, that's a good one; now pull the other one, it has bells on it,” into something that, these days, is extremely viable. How'd that happen?John: So, [sigh] I think that's just a—again, back to trends—I think that's just a trend that we're seeing from cloud providers and listening to their customers and continuing to improve the service. So, one of the reasons that HPC was—especially we'll call it capacity-level HPC or large HPC, right—you've always been able to run high throughput; the cloud is a high throughput machine, right? You can run a thousand disconnected VMs no problem, auto-scaling, anybody who runs a massive web front-end can attest to that. But what we saw with HPC—and we used to call those [grid 00:12:45] jobs, right, the small, decoupled computing jobs—but what we've seen is a huge increase in the quality of the underlying fabric—things like RDMA being made available, things like improved network locality, where you now have predictive latency between your nodes or between your VMs—and I think those, combined with the huge investment that companies like AWS have made in their file systems, the huge investment companies like Google have made in their data storage systems have made HPC viable, especially at a small-scale—for cloud-based HPC specifically—viable for organizations.And for a small engineering team, who's looking to run say, computer-aided engineering simulation or who's looking to prototype some new way of testing or doing some kind of simulation, it's a huge, huge improvement in speed because now they don't have to order a dozen or two dozen or five dozen nodes, have them shipped, rack them, stack them, cool them, power them, right? They can just spin up the resource in the cloud, test it out, try their simulation, try out the new—the software that they want, and then spin it all down if it doesn't work. So, that elasticity has also been huge. And again, I think the big—to kind of summarize, I think the big driver there is the improvement in this the service itself, right? We're seeing cloud providers taking that discipline a little bit more seriously.Corey: I still see that there are cases where the raw math doesn't necessarily add up for sustained, long-term use cases. But I also see increasingly that with HPC, that's usually not what the workload looks like. With, you know, the exception of we're going to spend the next 18 months training some new LLM thing, but even then the pricing is ridiculous. What is it their new P6 or whatever it is—P5—the instances that have those giant half-rack Nvidia cards that are $800,000 and so a year each if you were to just rent them straight out, and then people running fleets of these things, it's… wow that's more commas in that training job than I would have expected. But I can see just now the availability for driving some of that, but the economics of that once you can get them in your data center doesn't strike me as being particularly favoring the cloud.John: Yeah, there's a couple of different reasons. So, it's almost like an inverse curve, right? There's a crossover point or a breakeven point at which—you know, and you can make this argument with almost any level of infrastructure—if you can keep it sufficiently full, whether it's AI training, AI inference, or even traditional HPC if you can keep the machine or the group of machines sufficiently full, it's probably cheaper to buy it and put it in your facility. But if you don't have a facility or if you don't need to use it a hundred percent of the time, the dividends aren't always there, right? It's not always worth, you know, buying a $250,000 compute system, you know, like say, an Nvidia, as you—you know, like, a DGX, right, is a good example.The DGX H100, I think those are a couple $100,000. If you can't keep that thing full and you just need it for training jobs or for development and you have a small team of developers that are only going to use it six hours a day, it may make sense to spin that up in the cloud and pay for a fractional use, right? It's no different than what HPC has been doing for probably the past 50 years with national supercomputing centers, which is where my background came from before cloud, right? It's just a different model, right? One is public economies of, you know, insert your credit card and spend as much as you want and the other is grant-funded and supporting academic research, but the economy of scales is kind of the same on both fronts.Corey: I'm also seeing a trend that this is something that is sort of disturbing when you realize what I've been doing and how I've been going about things, that for the last couple of years, people actually started to care about the AWS bill. And I have to say, I felt like I was severely out of sync with a lot of the world the first few years because there's giant savings lurking in your AWS bill, and the company answer in many cases was, “We don't care. We'd rather focus our energies on shipping faster, building something new, expanding, capturing market.” And that is logical. But suddenly those chickens are coming home to roost in a big way. Our phone is ringing off the hook, as I'm sure you've noticed and your time here, and suddenly money means something again. What do you think drove it?John: So, I think there's a couple of driving factors. The first is obviously the broader economic conditions, you know, with the economic growth in the US, especially slowing down post-pandemic, we're seeing organizations looking for opportunities to spend less to be able to deliver—you know, recoup that money and deliver additional value. But beyond that, right—because, okay, but startups are probably still lighting giant piles of VC money on fire, and that's okay, but what's happening, I think, is that the first wave of CIOs that said cloud-first, cloud-only basically got their comeuppance. And, you know, these enterprises saw their explosive cloud bills and they saw that, oh, you know, we moved 5000 servers to AWS or GCP or Azure and we got the bill, and that's not sustainable. And so, we see a lot of cloud repatriation, cloud optimization, right, a lot of second-gen… cloud, I'll call them second-gen cloud-native CIOs coming into these large organizations where their predecessor made some bad financial decisions and either left or got asked to leave, and now they're trying to stop from lighting their giant piles of cash on fire, they're trying to stop spending 3X what they were spending on-prem.Corey: I think an easy mistake for folks to make is to get lost in the raw infrastructure cost. I'm not saying it's not important. Obviously not, but you could save a giant pile of money on your RDS instances by running your own database software on top of EC2, but I don't generally recommend folks do it because you also need engineering time to be focusing on getting those things up, care and feeding, et cetera. And what people lose sight of is the fact that the payroll expense is almost universally more than the cloud bill at every company I've ever talked to.So, there's a consistent series of, “Well, we're just trying to get to be the absolute lowest dollar figure total.” It's the wrong thing to emphasize on, otherwise, “Cool, turn everything off and your bill drops to zero.” Or, “Migrate to another cloud provider. AWS bill becomes zero. Our job is done.” It doesn't actually solve the problem at all. It's about what's right for the business, not about getting the absolute lowest possible score like it's some kind of code golf tournament.John: Right. So, I think that there's a couple of different ways to look at that. One is obviously looking at making your workloads more cloud-native. I know that's a stupid buzzword to some people, but—Corey: The problem I have with the term is that it means so many different things to different people.John: Right. But I think the gist of that is taking advantage of what the cloud is good at. And so, what we saw was that excess capacity on-prem was effectively free once you bought it, right? There were there was no accountability for burning through extra V CPUs or extra RAM. And then you had—Corey: Right. You spin something up in your data center and the question is, “Is the physical capacity there?” And very few companies had a reaping process until they were suddenly seeing capacity issues and suddenly everyone starts asking you a whole bunch of questions about it. But that was a natural forcing function that existed. Now, S3 has infinite storage, or it might as well. They can add capacity faster than you can fill it—I know this; I've tried—and the problem that you have then is that it's always just a couple more cents per gigabyte and it keeps on going forever. There's no, we need to make an investment decision because the SAN is at 80% capacity. Do you need all those 16 copies of the production data that you haven't touched since 2012? No, I probably don't.John: Yeah, there's definitely a forcing function when you're doing your own capacity planning. And the cloud, for the most part, as you've alluded to, for most organizations is infinite capacity. So, when they're looking at AWS or they're looking at any of the public cloud providers, it's a potentially infinite bill. Now, that scares a lot of organizations, and so because they didn't have the forcing function of, hey, we're out of CPUs, or we're out of hard disk space, or we're out of network ports, I think that because the cloud was a buzzword that a lot of shareholders and boards wanted to see in IT status reports and IT strategic plans, I think we grew a little bit further than we should have, from an enterprise perspective. And I think a lot of that's now being clawed back as organizations are maturing and looking to manage cost. Obviously, the huge growth of just the term FinOps from a search perspective over the last three years has cemented that, right? We're seeing a much more cost-conscious consumer—cloud consumer—than we saw three years ago.Corey: I think that the baseline level of understanding has also risen. It used to be that I would go into a client environment, prepared to deploy all kinds of radical stuff that these days look like context-aware architecture and things that would automatically turn down developer environments when developers were done for the day or whatnot. And I would discover that, oh, you haven't bought Reserved Instances in three years. Maybe start there with the easy thing. And now you don't see those, the big misconfigurations or the big oversights the way that you once did.People are getting better at this, which is a good thing. I'm certainly not having a problem with this. It means that we get to focus on things that are more architecturally nuanced, which I love. And I think that it forces us to continue innovating rather than just doing something that basically any random software stack could provide.John: Yeah, I think to your point, the easy wins are being exhausted or have been exhausted already, right? Very rarely do we walk into a customer and see that they haven't bought a, you know, Reserved Instance, or a Savings Plan. That's just not a thing. And the proliferation of software tools to help with those things, of course, in some cases, dubious proposition of, “We'll fix your cloud bill automatically for a small percentage of the savings,” that some of those software tools have, I think those have kind of run their course. And now you've got a smarter populace or smarter consumer and it does come into the more nuanced stuff, right.All right, do you really need to replicate data across AZs? Well, not if your workloads aren't stateful. Well, so some of the old things—and Kubernetes is a great example of this, right—the age old adage of, if I'm going to spin up an EKS cluster, I need to put it in three AZs, okay, why? That's going to cost you money [laugh], the cross-AZ traffic. And I know cross-AZ traffic is a simple one, but we still see that. We still see, “Well, I don't know why I put it across all three AZs.”And so, the service-to-service communication inside that cluster, the control plane traffic inside that cluster, is costing you money. Now, it might be minimal, but as you grow and as you scale your product or the services that you're providing internally, that may grow to a non-trivial sum of money.Corey: I think that there's a tipping point where an unbounded growth problem is always going to emerge as something that needs attention and needs to be focused on. But I should ask you this because you have a skill set that is, as you know, extremely in demand. You also have that rare gift that I wish wasn't as rare as it is where you can be thrown into the deep end knowing next to nothing about a particular technology stack, and in a remarkably short period of time, develop what can only be called subject matter expertise around it. I've seen you do this years past with Kubernetes, which is something I'm still trying to wrap my head around. You have a natural gift for it which meant that, from many respects, the world was your oyster. Why this? Why now?John: So, I think there's a couple of things that are unique at this thing, at this time point, right? So obviously, helping customers has always been something that's fun and exciting for me, right? Going to an organization and solving the same problem I've solved 20 different times, for example, spinning up a Kubernetes cluster, I guess I have a little bit of a little bit of squirrel syndrome, so to speak, and that gets—it gets boring. I'd rather just automate that or build some tooling and disseminate that to the customers and let them do that. So, the thing with cost management is, it's always a different problem.Yeah, we're solving fundamentally the same problem, which is, I'm spending too much, but it's always a different root cause, you know? In one customer, it could be data transfer fees. In another customer, it could be errant development growth where they're not controlling the spend on their development environments. In yet another customer, it could be excessive object storage growth. So, being able to hunt and look for those and play detective is really fun, and I think that's one of the things that drew me to this particular area.The other is just from a timing perspective, this is a problem a lot of organizations have, and I think it's underserved. I think that there are not enough companies—service providers, whatever—focusing on the hard problem of cost optimization. There's too many people who think it's a finance problem and not enough people who think it's an engineering problem. And so, I wanted to do work on a place where we think it's an engineering problem.Corey: It's been a very… long road. And I think that engineering problems and people problems are both fascinating to me, and the AWS bill is both. It's often misunderstood as a finance problem, and finance needs to be consulted absolutely, but they can't drive an optimization project, and they don't know what the context is behind an awful lot of decisions that get made. It really is breaking down bridges. But also, there's a lot of engineering in here, too. It scratches my itch in that direction, anyway.John: Yeah, it's one of the few business problems that I think touches multiple areas. As you said, it's obviously a people problem because we want to make sure that we are supporting and educating our staff. It's a process problem. Are we making costs visible to the organization? Are we making sure that there's proper chargeback and showback methodologies, et cetera? But it's also a technology problem. Did we build this thing to take advantage of the architecture or did we shoehorn it in a way that's going to cost us a small fortune? And I think it touches all three, which I think is unique.Corey: John, I really want to thank you for taking the time to speak with me. If people want to learn more about what you're up to in a given day, where's the best place for them to find you?John: Well, thanks, Corey, and thanks for having me. And, of course obviously, our website duckbillgroup.com is a great place to find out what we're working on, what we have coming. I also, I'm pretty active on LinkedIn. I know that's [laugh]—I'm not a huge Twitter guy, but I am pretty active on LinkedIn, so you can always drop me a follow on LinkedIn. And I'll try to post interesting and useful content there for our listeners.Corey: And we will, of course, put links to that in the [show notes 00:28:37], which in my case, is of course extremely self-aggrandizing. But that's all right. We're here to do self-promotion. Thank you so much for taking the time to chat with me, John. I appreciate it. Now, get back to work.John: [laugh]. All right, thanks, Corey. Have a good one.Corey: John Wynkoop, cloud economist at The Duckbill Group. I'm Cloud Economist Corey Quinn, and this is Screaming in the Cloud. If you've enjoyed this podcast, please leave a five-star review on your podcast platform of choice, whereas if you've hated this podcast, please leave a five-star review on your podcast platform of choice while also taking pains to note how you're using multiple podcast platforms these days because that just seems to be the way the world went.Corey: If your AWS bill keeps rising and your blood pressure is doing the same, then you need The Duckbill Group. We help companies fix their AWS bill by making it smaller and less horrifying. The Duckbill Group works for you, not AWS. We tailor recommendations to your business and we get to the point. Visit duckbillgroup.com to get started.