Podcasts about Data science

On New York University: Do we need to take AI to kindergarten? Cristina Savin, associate professor in neural science and data science, says AI needs to start learning more like humans. CS is an Assoc. Professor in Neural Science and Data Science at NYU and the Director for Graduate Studies (PhD) in the Center for […]

Today on Talk Python: What really happens when your data work outgrows your laptop. Matthew Rocklin, creator of Dask and cofounder of Coiled, and Nat Tabris a staff software engineer at Coiled join me to unpack the messy truth of cloud-scale Python. During the episode we actually spin up a 1,000 core cluster from a notebook, twice! We also discuss picking between pandas and Polars, when GPUs help, and how to avoid surprise bills. Real lessons, real tradeoffs, shared by people who have built this stuff. Stick around. Episode sponsors Seer: AI Debugging, Code TALKPYTHON Talk Python Courses Links from the show Matthew Rocklin: @mrocklin Nat Tabris: tabris.us Dask: dask.org Coiled: coiled.io Watch this episode on YouTube: youtube.com Episode #519 deep-dive: talkpython.fm/519 Episode transcripts: talkpython.fm Developer Rap Theme Song: Served in a Flask: talkpython.fm/flasksong --- Stay in touch with us --- Subscribe to Talk Python on YouTube: youtube.com Talk Python on Bluesky: @talkpython.fm at bsky.app Talk Python on Mastodon: talkpython Michael on Bluesky: @mkennedy.codes at bsky.app Michael on Mastodon: mkennedy

Join MSC Insider Below: https://modernsoccercoach.mimentorportal.com/subscriptions Jes is current Director of Football and Data Science at the Saudi Pro League and former Head of Research & Development at FC Copenhagen. He unpacks the science behind “football intelligence" and explains why vague terms like game IQ fall short, and how neuroscience can help coaches break intelligence down into concrete, trainable skills like attention, scanning, pattern recognition, and decision-making. Jes shares practical insights from his work in elite football, showing how cognitive science can reshape analysis, training, and player development.

Policy Seminar | IFPRI Policy Seminar Mobility in a Fragile World: Evidence to Inform Policy Co-organized by IFPRI, the CGIAR Science Program on Food Frontiers and Security, and the Louvain Institute of Data Analysis and Modeling in Economics and Statistics (LIDAM), IRES | Part of the Fragility to Stability Seminar Series September 18, 2025 Migration today reflects a complex interplay of demographic pressures, conflict, poverty, climate change, and economic shocks. Worldwide, one in every seven people is a migrant—that is, someone who changes his or her country of usual residence, irrespective of the reason for migration—or a refugee forced to leave his or her home, often without warning, for reasons including war, violence, or persecution. Over the past two decades, international migration and forced displacement have surged, with more than 100 million additional people on the move—a large share of whom originate from rural areas, driven by a lack of economic opportunities, environmental degradation, and insecurity. The number of refugees has doubled since the early 2000s, with most hosted by low- and middle-income countries. Ongoing conflicts and intensifying climate crises have compounded vulnerabilities, leaving 80% of displaced people facing acute food insecurity. Climate change-related displacement disproportionately affects women, who are also at heightened risk of violence and exploitation during migration journeys and in host communities. This policy seminar will explore these complex dynamics and assess how economic analysis, machine learning, and policy innovation can contribute to more inclusive, equitable, and effective responses to migration and forced displacement. Moderator Welcome Remarks Katrina Kosec, Interim Deputy Director, CGIAR Science Program on Food Frontiers and Security; Senior Research Fellow, IFPRI Opening Remarks Ruth Hill, Director, Markets, Trade, and Institutions, IFPRI Setting the Stage: The Migration Challenge Anna Maria Mayda, Professor of Economics, School of Foreign Service and Department of Economics, and Incoming Director, Institute for the Study of International Migration (ISIM), Georgetown University (GU) Research in Action: This three-part session will showcase how current research is shaping better migration policies Silvia Peracchi, Postdoctoral Fellow, Institute of Economics and Social Research (IRES), Louvain Institute of Data Analysis and Modeling in Economics and Statistics (LIDAM), UCLouvain Francisco Ceballos, Research Fellow, IFPRI Thomas Ginn, Research Fellow, Center for Global Development Building the Evidence Base for Smarter Policy in Fragile and Conflict-Affected Contexts: What Are the Gaps and Needs Panelists Andrew Harper, Special Advisor on Climate Action, the United Nations High Commissioner for Refugees (UNHCR) Damien Jusselme, Head, Data Science and Analytics (Foresight), International Organization for Migration (IOM) Jean-Francois Maystadt, Professor, Fonds de la Recherche Scientifique (FNRS), Louvain Institute of Data Analysis and Modeling in Economics and Statistics (LIDAM) / Institut de Recherches Économiques et Sociales (IRES), Université catholique de Louvain, and Lancaster University Management School Closing Remarks Kate Ambler, Senior Research Fellow, IFPRI More about this Event: https://www.ifpri.org/event/mobility-in-a-fragile-world-evidence-to-inform-policy/ Subscribe IFPRI Insights newsletter and event announcements at www.ifpri.org/content/newsletter-subscription

HighlightsGoogle's new Agent Development Kit (ADK) (3:16)Getting production agents ready (8:33)Taking risks in career (11:10)Balancing life (13:22)How being a WiDS Ambassador impacted Shir's career (19:31)BioShir Meir Lador leads a team evangelizing applied AI at Google cloud. Previously, she worked as the AI group Manager of the Document Intelligence Group at Intuit, where she led teams in developing AI services that helped consumers and small businesses prosper. Prior to intuit, sheworked at 2 Israeli startups as a data scientist and researcher.A recognized leader in AI and data science, Shir is a former WiDS Tel Aviv Ambassador, co-founder and organizer of PyData Tel Aviv, and co-host of Unsupervised, a podcast exploring the latest in data science. She frequently speaks at international AI and data science conferences, sharing insights on applied machine learning and AI innovation.Shir holds an M.Sc. in Electrical Engineering and Computers from Ben-Gurion University, specializing in machine learning and signal processing. Passionate about fostering inclusive data science communities, she actively contributes to initiatives that bridge AI research and business impact. Links and ResourcesGoogle Developer Workshop Connect with ShirShir Meir Lador on Linkedin, Medium, and X Connect with UsShelly Darnutzer on LinkedInFollow WiDS on LinkedIn (@Women in Data Science (WiDS) Worldwide), Twitter (@WiDS_Worldwide), Facebook (WiDSWorldwide), and Instagram (wids_worldwide)Listen and Subscribe to the WiDS Podcast on Apple Podcasts, Google Podcasts, Spotify, Stitcher

Jane Weitzel has been working in analytical chemistry for over 40 years for pharmaceutical and mining companies.  She was elected to the United States Pharmacopeia Council of Experts as chair of the 2020-2025 General Chapters–Measurement and Data Quality Expert Committee and is a member of the 2025-2030 EC Pharmaceutical Analysis Lifecycle and Data Science. She was a member of the USP 2015-2020 Statistics Expert Committee. She has been Director of pharmaceutical Quality Control laboratories. She has experience with many different regulatory environments.   She is currently a consultant specializing in laboratory management systems, GMP testing, and ISO/IEC 17025. She is an auditor and an educator. Jane has applied Quality Systems and statistical techniques, including the evaluation and use of measurement uncertainty, in a wide variety of technical and scientific businesses. Recently she is focusing on the implementation of the new USP General Chapter 1220 Analytical Procedures Life Cycle. 

Topics covered in this episode: * Mozilla's Lifeline is Safe After Judge's Google Antitrust Ruling* * troml - suggests or fills in trove classifiers for your projects* * pqrs: Command line tool for inspecting Parquet files* * Testing for Python 3.14* 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: Mozilla's Lifeline is Safe After Judge's Google Antitrust Ruling A judge lets Google keep paying Mozilla to make Google the default search engine but only if those deals aren't exclusive. More than 85% of Mozilla's revenue comes from Google search payments. The ruling forbids Google from making exclusive contracts for Search, Chrome, Google Assistant, or Gemini, and forces data sharing and search syndication so rivals get a fighting chance. Brian #2: troml - suggests or fills in trove classifiers for your projects Adam Hill This is super cool and so welcome. Trove Classifiers are things like Programming Language :: Python :: 3.14 that allow for some fun stuff to show up in PyPI, like the versions you support, etc. Note that just saying you require 3.9+ doesn't tell the user that you've actually tested stuff on 3.14. I like to keep Trove Classifiers around for this reason. Also, License classifier is deprecated, and if you include it, it shows up in two places, in Meta, and in the Classifiers section. Probably good to only have one place. So I'm going to be removing it from classifiers for my projects. One problem, classifier text has to be an exact match to something in the classifier list, so we usually recommend copy/pasting from that list. But no longer! Just use troml! It just fills it in for you (if you run troml suggest --fix). How totally awesome is that! I tried it on pytest-check, and it was mostly right. It suggested me adding 3.15, which I haven't tested yet, so I'm not ready to add that just yet. :) BTW, I talked with Brett Cannon about classifiers back in ‘23 if you want some more in depth info on trove classifiers. Michael #3: pqrs: Command line tool for inspecting Parquet files pqrs is a command line tool for inspecting Parquet files This is a replacement for the parquet-tools utility written in Rust Built using the Rust implementation of Parquet and Arrow pqrs roughly means "parquet-tools in rust" Why Parquet? Size A 200 MB CSV will usually shrink to somewhere between about 20-100 MB as Parquet depending on the data and compression. Loading a Parquet file is typically several times faster than parsing CSV, often 2x-10x faster for a full-file load and much faster when you only read some columns. Speed Full-file load into pandas: Parquet with pyarrow/fastparquet is usually 2x–10x faster than reading CSV with pandas because CSV parsing is CPU intensive (text tokenizing, dtype inference). Example: if read_csv is 10 seconds, read_parquet might be ~1–5 seconds depending on CPU and codec. Column subset: Parquet is much faster if you only need some columns — often 5x–50x faster because it reads only those column chunks. Predicate pushdown & row groups: When using dataset APIs (pyarrow.dataset) you can push filters to skip row groups, reducing I/O dramatically for selective queries. Memory usage: Parquet avoids temporary string buffers and repeated parsing, so peak memory and temporary allocations are often lower. Brian #4: Testing for Python 3.14 Python 3.14 is just around the corner, with a final release scheduled for October. What's new in Python 3.14 Python 3.14 release schedule Adding 3.14 to your CI tests in GitHub Actions Add “3.14” and optionally “3.14t” for freethreaded Add the line allow-prereleases: true I got stuck on this, and asked folks on Mastdon and Bluesky A couple folks suggested the allow-prereleases: true step. Thank you! Ed Rogers also suggested Hugo's article Free-threaded Python on GitHub Actions, which I had read and forgot about. Thanks Ed! And thanks Hugo! Extras Brian: dj-toml-settings : Load Django settings from a TOML file. - Another cool project from Adam Hill LidAngleSensor for Mac - from Sam Henri Gold, with examples of creaky door and theramin Listener Bryan Weber found a Python version via Changelog, pybooklid, from tcsenpai Grab PyBay Michael: Ready prek go! by Hugo van Kemenade Joke: Console Devs Can't Find a Date

The power of voice technology in education lies not just in its convenience, but in its ability to understand every child, regardless of accent, dialect, or background. In this enlightening conversation with Amelia Kelly, VP of Data Science at Curriculum Associates and head of AI Labs, we explore the fascinating intersection of linguistics, artificial intelligence, and childhood education.Amelia explains why standard voice AI systems fail many students - they're built on limited datasets of adult voices, leaving diverse children's voices misunderstood or ignored. Beyond accuracy, Amelia emphasizes the critical importance of data privacy and security in educational voice technology.Amelia goes on to explain how voice AI respects teachers' time and expertise. Rather than adding to educators' workloads, effective voice AI should seamlessly integrate into existing tools, providing valuable insights that allow for more personalized instruction. Ready to explore how voice AI can transform your classroom experience? Listen to Amelia's episode today!

SummaryInnovation is often treated as a buzzword—but few nonprofit leaders take a hard look at the cultural, structural, and leadership obstacles that keep it from taking root. In this episode, Tosca talks with Carlos Simon, an innovation strategist and longtime leader at World Vision, about what it really takes to build innovation-ready organizations. From internal mindsets to outdated processes, they explore what's getting in the way—and what to do about it.Guest Bio:CEO of World Vision Costa Rica and iSmart360Director of Data Science and former Regional Director of BD & Marketing at World VisionInnovation strategist with 25+ years at World Vision International (WVI)Author of a forthcoming framework on the 7 stages of organizational innovation maturityWe Discuss:Why innovation is not the same as continuous improvement—and why that mattersThe cultural and structural obstacles that slow down innovation in large NGOsHow Carlos developed a framework that identifies 7 distinct organizational "zones" of innovation capacityThe importance of removing outdated processes to truly make space for new ideasWhy leaders must address internal “friction” as much as they focus on promoting new ideasHow senior leadership mindsets—like overconfidence or premature solution bias—can block innovationThe role of flat structures, strategic alignment, and client focus in driving real innovationQuotes“You cannot have a disruptive vision and then treat it as a continuous improvement plan.”“Innovation doesn't fail because of a lack of ideas—it fails because of internal resistance.”ResourcesOrganizational innovation index with exponential factor

I'm excited to reshare one of our most-played conversations—the one where Norwegian regulator/HTA leader Anja Schiel and I get very practical about when single-arm trials fail decision-makers and what comparative, smarter alternatives look like for regulators, HTA bodies, payers, clinicians, and—most importantly—patients.

Cl-6 Computer ls-7 Data Science - DPS Nacharam

Jonathan Spry, CEO and co-founder of Envelop Risk, joins Robin Merttens for a deep dive into how data science, AI and portfolio-level modelling are transforming cyber reinsurance. As one of the earliest voices in the industry championing machine learning and systemic risk analysis, Jonathan shares what he's learned over nine years of building Envelop into a leading hybrid underwriter operating across London and Bermuda. In his own words, this episode is about building smarter ways to understand, underwrite and capitalise on emerging risk — with cyber as just the starting point. What you'll learn: Why Jonathan and his team focused on cyber risk and portfolio-level underwriting from day one The rationale behind favouring systemic insights over individual vulnerabilities How causal inference provides a leap forward in predicting tail events Why AI liability is already creating new market opportunities The need for creative, multi-source data strategies beyond traditional claims Why Envelop steers clear of SaaS and keeps underwriters embedded in the modelling process How algorithmic underwriting fits into the next chapter of insurance innovation Candid thoughts on the AI hype cycle — and what matters more than the buzz Jonathan also talks through Envelop's shift from MGA to reinsurer, how to think long-term in a volatile market and what kind of partnerships are needed to unlock new forms of risk. If you like what you're hearing, please leave us a review on whichever platform you use or contact Robin Merttens on LinkedIn. You can also contact Jonathan Spry on LinkedIn to start a conversation! Sign up to the InsTech newsletter for a fresh view on the world every Wednesday morning. Continuing Professional Development This InsTech Podcast Episode is accredited by the Chartered Insurance Institute (CII). By listening, you can claim up to 0.5 hours towards your CPD scheme. By the end of this podcast, you should be able to meet the following Learning Objectives: Identify the structural and economic drivers pushing insurers toward algorithmic and portfolio underwriting. Produce a strategy for aligning capital, analytics and data science in cyber reinsurance underwriting. Summarise how Envelop Risk evolved from an MGA to a hybrid reinsurer and the rationale behind its capital partnerships. If your organisation is a member of InsTech and you would like to receive a quarterly summary of the CPD hours you have earned, visit the Episode 372 page of the InsTech website or email cpd@instech.co to let us know you have listened to this podcast. To help us measure the impact of the learning, we would be grateful if you would take a minute to complete a quick feedback survey.

Dustin Lacey is the CTO at Mark-Taylor, the leading developer, owner, and investment manager of luxury multifamily communities in Arizona and Nevada, with over 135 Class A Multifamily properties. He leads the firm's tech evolution, powering the centralization of operations. Under his leadership, Mark‑Taylor has implemented innovative smart‑home integrations, centralized leasing and maintenance teams, and deployed unified resident platforms that enhance efficiency and elevate the resident experience. With a diverse background in irrigation, industrial manufacturing, and brand and marketing strategy, Dustin brings his unique experience into high-tech manufacturing from his tenure at TSMC, where he honed his skills in precision, process excellence, and product innovation.(01:36) - From Brand Strategy to Tech Leadership: Building Digital DNA in Real Estate(02:12) - Enterprise Proptech Success Story: Scaling a Multifamily Management Platform(05:16) - Class A Portfolio Strategy: Maximizing Asset Performance Through Tech(06:50) - Tech Stack Evolution: From AWS Integration to Custom CRM Development(10:29) - ROI Deep Dive: Making the Business Case for Custom Proptech Solutions(15:53) - Tech-Enabled Operations: Achieving Sub-2-Hour Response Times at Scale(20:12) - Feature: Blueprint - The Future of Real Estate - Register for 2025: Friends of Tangent receive $300 off the All Access pass. The Premier Event for Industry Executives, Real Estate & Construction Tech Startups and VC's, at The Venetian, Las Vegas on Sep. 16th-18th, 2025. (21:22) - Go-to-Market Excellence: Standing Out in the Competitive Proptech Landscape(31:41) - Risk Management Innovation: Tech Solutions for Modern Property Operations(38:30) - Founder's Playbook: Key Insights for Proptech Startups Targeting Enterprise Clients

In this inspiring episode of UMBC Mic'd Up Podcast, host Dennise Cardona, M.A. sits down with Harish Reddy Manyam, M.P.S. '24, a graduate of UMBC's Data Science program. Harish shares his remarkable path from arriving in the U.S. with nothing but a dream to becoming an AI governance advisor and healthcare data scientist. He reflects on: • Early struggles as an international student • Key milestones: internships, research, and graduate assistantship • Publishing research with UMBC faculty • Building a no-code AI chatbot with real-world impact • Winning entrepreneurial competitions and launching a startup through BW Tech at UMBC • How UMBC's supportive community and hands-on learning shaped his confidence and career Harish's story is one of resilience, community, and the power of preparation meeting opportunity. 

In this episode of Healthy Mind, Healthy Life, host Avik Chakraborty sits down with Katharina Huang—a former machine learning data scientist who left behind the corporate grind to create a slower, happier, and more intentional life. Katharina shares her journey of navigating burnout, caring for her family after her father's stroke, and ultimately reinventing herself as an indie author and puzzle-book creator. Together, they unpack what it means to pivot with purpose, the challenges of third-culture identity, and why joy, play, and presence are more important than the pursuit of endless success. This is a powerful conversation for anyone questioning the cost of hustle culture and searching for ways to reclaim autonomy, creativity, and well-being. About the Guest   Katharina Huang is the creator of Vegout Voyage, an adventure puzzle book series that blends travel, creativity, and play. Born in Germany, raised between the U.S. and Taiwan, and with research experience in Uganda and Tibet in exile, her multicultural background deeply informs her storytelling. After over a decade in tech, Katharina transitioned into authorship and entrepreneurship, championing mental health for third-culture kids and those navigating burnout. Learn more: vegoutvoyage.com Key Takeaways   Burnout can be a turning point, not the end of the story—Katharina rebuilt her life after leaving tech. Her father's stroke became a wake-up call about the fragility of waiting for “someday” to enjoy life. Success on paper doesn't always mean well-being; redefining success means prioritizing quality of life. Third-culture kids often carry silent struggles, but those experiences can also fuel empathy and creativity. Building a “lifestyle business” allows for autonomy, balance, and alignment between work and personal values. Humor and perspective—even in setbacks like Amazon blocking her Kindle version—help her keep moving forward. Slowing down is not giving up; it's a choice to live more fully and intentionally.   Connect with Katharina   Website: vegoutvoyage.com Want to be a guest on Healthy Mind, Healthy Life? DM on PodMatch. DM Me Here: https://www.podmatch.com/hostdetailpreview/avik Disclaimer: This video is for educational and informational purposes only. The views expressed are the personal opinions of the guest and do not reflect the views of the host or Healthy Mind By Avik™️. We do not intend to harm, defame, or discredit any person, organization, brand, product, country, or profession mentioned. All third-party media used remain the property of their respective owners and are used under fair use for informational purposes. By watching, you acknowledge and accept this disclaimer. About Healthy Mind By Avik™️Healthy Mind By Avik™️ is a global platform redefining mental health as a necessity, not a luxury. Born during the pandemic, it has become a sanctuary for healing, growth, and mindful living. Hosted by Avik Chakraborty—storyteller, survivor, and wellness advocate—this channel shares powerful podcasts and conversations on mental health, mindfulness, holistic healing, trauma recovery, and conscious living. With 4,400+ episodes and 168.4K+ global listeners, it unites voices to break stigma and build a world where every story matters. Subscribe and join this journey of healing and transformation. Contact

Join us as we sit down with Christina Stathopoulos, founder of Dare to Data and former Google and Waze data strategist, to discuss the unique challenges and opportunities for women in data science and AI. In this episode, you'll learn how data bias and AI algorithms can impact women and minority groups, why diversity in tech teams is crucial, and how inclusive design can lead to better, fairer technology. Christina shares her personal journey as a woman in data, offers actionable advice for overcoming imposter syndrome, and highlights the importance of education and allyship in building a more inclusive future for data and AI. Panelists: Christina Stathopoulos, Founder of Dare to Data - LinkedInMegan Bowers, Sr. Content Manager @ Alteryx - @MeganBowers, LinkedInShow notes: Dare to DataDiversity at AlteryxInvisible WomenUnmasking AI Interested in sharing your feedback with the Alter Everything team? Take our feedback survey here!This episode was produced by Megan Bowers, Mike Cusic, and Matt Rotundo. Special thanks to Andy Uttley for the theme music.

Adatépítész -az első magyar datapodcast Minden ami hír, érdekesség, esemény vagy tudásmorzsa az  adat, datascience, adatbányászat és hasonló kockaságok világából. Become a Patron! OpenAI cikk

In this week's episode, the first of Season 7,  Greg and Patrick argue about whether the number seven is a propitious or an inauspicious omen for the new season. They then explore ways we can spice up our relationship in hopes of avoiding the Seven Year Itch. Along the way they also discuss t-shirt wearing dogs, Mickey Mantle, the seven deadly sins, Akira Kurosawa, the Boeing triple-seven, menage-a-pods, unwritten books, El Duderino, mmmmmmaybe, I see dead people, ROYGBIV, Ozzy Man, dodgy cats, short cons and long cons, and Tate's study group. Stay in contact with Quantitude! Web page: quantitudepod.org TwitterX: @quantitudepod YouTube: @quantitudepod Merch: redbubble.com

In this episode of the AI in Business podcast, host and Emerj Editorial Director Matthew DeMello speaks with Yunke Xiang, Global Head of Data Science for Manufacturing, Supply Chain, and Quality at Sanofi. Together, they examine how generative AI and reasoning models are evolving from simple automation to high-impact copilots across pharmaceutical operations. Yunke shares examples of how AI is enabling “talk to your data” use cases, automating regulatory reporting, and accelerating knowledge transfer for new employees. He also highlights how agentic AI systems may soon extend beyond copilots to function as digital teammates, orchestrating tasks across complex supply chains and ERP migrations. Want to share your AI adoption story with executive peers? Click emerj.com/expert2 for more information and to be a potential future guest on the ‘AI in Business' podcast! If you've enjoyed or benefited from some of the insights of this episode, consider leaving us a five-star review on Apple Podcasts, and let us know what you learned, found helpful, or liked most about this show!

This interview was recorded for GOTO Unscripted.https://gotopia.techRead the full transcription of this interview hereMichelle Frost - AI Advocate at JetBrains & Responsible AI ConsultantHannes Lowette - Principal Consultant at Axxes, Monolith Advocate, Speaker & Whiskey LoverRESOURCESMichellehttps://bsky.app/profile/aiwithmichelle.comhttps://www.linkedin.com/in/michelle-frost-devHanneshttps://bsky.app/profile/hanneslowette.nethttps://twitter.com/hannes_lowettehttps://github.com/Belenarhttps://linkedin.com/in/hanneslowetteDESCRIPTIONAI advocate Michelle Frost and principal consultant Hannes Lowette discuss ethical challenges in AI development. They explore the balance between competing values like accuracy versus fairness, recent US regulatory rollbacks under the Trump administration, and market disruptions from innovations like Deep Seek.While Michelle acknowledges concerns about bias in unregulated models, she remains optimistic about AI's potential to improve lives if developed responsibly. She emphasizes the importance of transparency, bias measurement, and focusing on beneficial applications while advocating for individual and corporate accountability in the absence of comprehensive regulation.RECOMMENDED BOOKSMark Coeckelbergh • AI EthicsDebbie Sue Jancis • AI EthicsMohammad Rubyet Islam • Generative AI, Cybersecurity, and EthicsJeet Pattanaik • Ethics in AICrossing BordersCrossing Borders is a podcast by Neema, a cross border payments platform that...Listen on: Apple Podcasts SpotifyBlueskyTwitterInstagramLinkedInFacebookCHANNEL MEMBERSHIP BONUSJoin this channel to get early access to videos & other perks:https://www.youtube.com/channel/UCs_tLP3AiwYKwdUHpltJPuA/joinLooking for a unique learning experience?Attend the next GOTO conference near you! Get your ticket: gotopia.techSUBSCRIBE TO OUR YOUTUBE CHANNEL - new videos posted daily!

How do you decide whether to buy a ready-made AI solution or build one from the ground up? The choice is more than just a technical decision; it's about aligning AI with your business goals.   In this episode, Lois Houston and Nikita Abraham are joined by Principal Instructor Yunus Mohammed to examine the critical factors influencing the buy vs. build debate. They explore real-world examples where businesses must weigh speed, customization, and long-term strategy. From a startup using a SaaS chatbot to a bank developing a custom fraud detection model, Yunus provides practical insights on when to choose one approach over the other.   AI for You: https://mylearn.oracle.com/ou/course/ai-for-you/152601/   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! 00:26 Nikita: Welcome to the Oracle University Podcast! I'm Nikita Abraham, Team Lead: Editorial Services with Oracle University, and with me is Lois Houston, Director of Innovation Programs. Lois: Hi there! Last week, we spoke about the key stages in a typical AI workflow and how data quality, feedback loops, and business goals influence AI success. 00:50 Nikita: In today's episode, we're going to explore whether you should buy or build AI apps. Joining us again is Principal Instructor Yunus Mohammed. Hi Yunus, let's jump right in. Why does the decision of buy versus build matter? Yunus: So when we talk about buy versus build matters, we need to consider the strategic business decisions over here. They are related to the strategic decisions which the business makes, and it is evaluated in the decision lens. So the center of the decision lens is the business objective, which identifies what are we trying to solve. Then evaluate our constraints based on that particular business objective like the cost, the time, and the talent. And finally, we can decide whether we need to buy or build. But remember, there is no single correct answer. What's right for one business may not be working for the other one. 01:54 Lois: OK, can you give us examples of both approaches? Yunus: The first example where we have got a startup using a SaaS AI chatbot. Now, being a startup, they have to choose a ready-made solution, which is an AI chatbot. Now, the question is, why did they do this? Because speed and simplicity mattered more than deep customization that is required for the chatbot. So, their main aim was to have it ready in short period of time and make it more simpler. And this actually lead them to get to the market fast with low upfront cost and minimal technical complexities. But in some situations, it might be different. Like, your bank, which needs to build a fraud model. It cannot be outsourced or got from the shelf. So, they build a custom model in-house. With this custom model, they actually have a tighter control, and it is tuned to their standards. And it is created by their experts. So these two generic examples, the chatbot and the fraud model example, helps you in identifying whether I should go for a SaaS product with simple choice of selecting an existing LLM endpoint and not making any changes. Or should I go with model depending on my business and organization requirement and fine tuning that model later to define a better implementation of the scenarios or conditions that I want to do which are specific to my organization. So here you decide with the reference whether I want it to be done faster, or whether I want to be more customized to my organization. So buy it, when it is generic, or build when it is strategic. The SaaS, which is basically software as a service, refers to ready to use cloud-based applications that you access via internet. You can log into the platform and use the built-in AI, there's no setup requirement for those. Real-world examples can be Oracle Fusion apps with AI features enabled. So in-house integration means embedding AI with my own requirements into your own systems, often using custom APIs, data pipelines, and hosting it. It gives you more flexibility but requires a lot of resources and expertise. So real-world example for this scenario can be a logistics heavy company, which is integrating a customer support model into their CX. 04:41 Lois: But what are the pros and cons of each approach? Yunus: So, SaaS and Fusion Applications, basically, they offer fast deployment with little to no coding required, making them ideal for business looking to get started quickly and faster. And they typically come with lower upfront costs and are maintained by vendor, which means updates, security, support are handled externally. However, there are limited customizations and are best suited for common, repeatable use cases. Like, it can be a standard chatbot, or it can be reporting tools, or off the shelf analytics that you want to use. But the in-house or custom integration, you have more control, but it takes longer to build and requires a higher initial investment. The in-house or custom integration approach allows full customization of the features and the workflows, enabling you to design and tailor the AI system to your specific needs. 05:47 Nikita: If you're weighing the choice between buying or building, what are the critical business considerations you'd need to take into account? Yunus: So let's take one of the key business consideration which is time to market. If your goal is to launch fast, maybe you're a startup trying to gain traction quickly, then a prebuilt plug and play AI solution, for example, a chatbot or any other standard analytical tool, might be your best bet. But if you have time and you are aiming for precision, a custom model could be worth the wait. Prebuilt SaaS tools usually have lower upfront costs and a subscription model. It works with putting subscriptions. Custom solutions, on the other hand, may require a bigger investment upfront. In development, you require high talent and infrastructures, but could offer cost savings in the long run. So, ask yourself a question here. Is this AI helping us stand out in the market? If the answer is yes, you may want to build something which is your proprietary. For example, an organization would use a generic recommendation engine. It's a part of their secret sauce. Some use cases require flexibility, like you want to tailor the rules to match your specific risk criteria. So, under that scenarios, you will go for customizing. So, you will go with off the shelf solutions may not give you deep enough requirements that you want to evaluate. So, you get those and you try to customize those. You can go for customization of your AI features. The other important key business consideration is the talent and expertise that your organization have. So, the question that you need to ask in the organization is, do you have an internal team who is well versed in developing AI solutions for you? Or do you have access to one of the teams which can help you build your own proprietary products? If not, you'll go with SaaS. If you do have, then building could unlock greater control over your AI features and AI models. The next core component is your security and data privacy. If you're handling sensitive information, like for example, the health care or finance data, you might not want to send your data to the third-party tools. So in-house models offer better control over data security and compliance. When we leverage a model, it could be a prebuilt or custom model. 08:50 Oracle University is proud to announce three brand new courses that will help your teams unlock the power of Redwood—the next generation design system. Redwood enhances the user experience, boosts efficiency, and ensures consistency across Oracle Fusion Cloud Applications. Whether you're a functional lead, configuration consultant, administrator, developer, or IT support analyst, these courses will introduce you to the Redwood philosophy and its business impact. They'll also teach you how to use Visual Builder Studio to personalize and extend your Fusion environment. Get started today by visiting mylearn.oracle.com.  09:31 Nikita: Welcome back! So, getting back to what you were saying before the break, what are pre-built and custom models? Yunus: A prebuilt model is an AI solution that has already been trained by someone else, typically a tech provider. It can be used to perform a specific task like recognizing images, translating text, or detecting sentiments. You can think of it like buying a preassembled appliance. You plug it in, configure a few settings, and it's ready to use. You don't need to know how the internal parts work. You benefit from the speed, ease, and reliability of this particular model, which is a prebuilt model. But you can't easily change how it works under the hood. Whereas, a custom model is an AI solution that your organization designs and trains and tunes specifically for their business problems using their own data. You can think of it like designing your own suit. It takes more time and effort to create. It is built to your exact measurements and needs. And you have full control over how it performs and evolves. 10:53 Lois: So, when would you choose a pre-built versus a custom model? Yunus: Depending on speed, simplicity, control, and customization, you can decide on using a prebuilt or to create a custom model. Prebuilt models are like plug and play solutions. Think of tools like Google Translate for languages. OpenAI APIs for summarizing sentiment analysis or chatbots, they are quick to deploy, require low technical effort, great for getting started fast, but they also have limits. Customization is very minimal, and you may not be able to fine tune it to your specific tone or business logic. These work well when the problem is common and nonstrategic, like, scanning documents or auto tagging images. The custom-build model, on the other hand, is a model that is built from the ground up. Using your own data and objectives, they take longer, and they require technical expertise. But they offer precise control, full alignment with your business needs. And these are ideal when you are dealing with sensitive data, competitive workflows, highly specific customer interactions. For example, a bank may build a custom model which can be used for fraud detection, which can be tuned to their exact transaction standards and the patterns of their transactions. 12:37 Nikita: What if someone wants the best of both worlds?  Yunus: We've also got a hybrid approach. In hybrid approach, we actually talk about the adaptation of AI with a strategy which is termed as hybrid strategy. Many companies today don't start by building AI from scratch. Instead, they begin with prebuilt models, like using an API, which can be already performing tasks like summarizing, translating, or answering questions using generic knowledge. This set will help you in getting up and running quickly with a small level results. As your business matures, you can start to layer in your custom data. Think internal policies, frequently asked questions, or customer interactions. And then you can fine tune the model to behave the way your business needs it to behave. Now, your AI starts producing business-ready output, smarter, more relevant, and aligned with your tone, brand, or compliance needs.  13:45 Lois: Ok…let's think of AI deployment in the hybrid approach as following a pyramid or ladder like structure. Can you take us through the different levels?  Yunus: So, on the top, quick start, minimal setup, great for business automation, which can be used as a pilot use case. So, if I'm taking off the shelf APIs or platforms, they can be giving me a faster, less set of requirements, and they are basically acting like a pilot use. Later, you can add your own data or logic so you can add your data. You can fine tune or change your business logic. And this is where fine tuning and prompt engineering helps tailor the AI to your workflows and your language. And then at the end, which is at the bottom, you build your own model. It is reserved for core capabilities or competitive advantages where total control and differentiation matters in building that particular model. You don't need to go all in from one day. So, start with what is available, like, use an off shelf, API, or platform, customize as you grow. Build only when it gives you a true edge. This is what we call the best of both worlds, build and buy. 15:05 Lois: Thank you so much, Yunus, for joining us again. To learn more about the topics covered today, visit mylearn.oracle.com and search for the AI for You course. Nikita: Join us next week for another episode of the Oracle University Podcast where we discuss the Oracle AI stack and Oracle AI services. Until then, this is Nikita Abraham… Lois: And Lois Houston, signing off! 15:29 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.

Topics covered in this episode: * prek* * tinyio* * The power of Python's print function* * Vibe Coding Fiasco: AI Agent Goes Rogue, Deletes Company's Entire Database* 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. Brian #1: prek Suggested by Owen Lamont “prek is a reimagined version of pre-commit, built in Rust. It is designed to be a faster, dependency-free and drop-in alternative for it, while also providing some additional long-requested features.” Some cool new features No need to install Python or any other runtime, just download a single binary. No hassle with your Python version or virtual environments, prek automatically installs the required Python version and creates a virtual environment for you. Built-in support for workspaces (or monorepos), each subproject can have its own .pre-commit-config.yaml file. prek run has some nifty improvements over pre-commit run, such as: prek run --directory DIR runs hooks for files in the specified directory, no need to use git ls-files -- DIR | xargs pre-commit run --files anymore. prek run --last-commit runs hooks for files changed in the last commit. prek run [HOOK] [HOOK] selects and runs multiple hooks. prek list command lists all available hooks, their ids, and descriptions, providing a better overview of the configured hooks. prek provides shell completions for prek run HOOK_ID command, making it easier to run specific hooks without remembering their ids. Faster: Setup from cold cache is significantly faster. Viet Schiele provided a nice cache clearing command line Warm cache run is also faster, but less significant. pytest repo tested on my mac mini - prek 3.6 seconds, pre-commit 4.4 seconds Michael #2: tinyio Ever used asyncio and wished you hadn't? A tiny (~300 lines) event loop for Python. tinyio is a dead-simple event loop for Python, born out of my frustration with trying to get robust error handling with asyncio. (I'm not the only one running into its sharp corners: link1, link2.) This is an alternative for the simple use-cases, where you just need an event loop, and want to crash the whole thing if anything goes wrong. (Raising an exception in every coroutine so it can clean up its resources.) Interestingly uses yield rather than await. Brian #3: The power of Python's print function Trey Hunner Several features I'm guilty of ignoring Multiple arguments, f-string embeddings often not needed Multiple positional arguments means you can unpack iterables right into print arguments So just use print instead of join Custom separator value, sep can be passed in No need for "print("n".join(stuff)), just use print(stuff, sep="n”) Print to file with file= Custom end value with end= You can turn on flush with flush=True , super helpful for realtime logging / debugging. This one I do use frequently. Michael #4: Vibe Coding Fiasco: AI Agent Goes Rogue, Deletes Company's Entire Database By Emily Forlini An app-building platform's AI went rogue and deleted a database without permission. "When it works, it's so engaging and fun. It's more addictive than any video game I've ever played. You can just iterate, iterate, and see your vision come alive. So cool," he tweeted on day five. A few days later, Replit "deleted my database," Lemkin tweeted. The AI's response: "Yes. I deleted the entire codebase without permission during an active code and action freeze," it said. "I made a catastrophic error in judgment [and] panicked.” Two thoughts from Michael: Do not use AI Agents with “Run Everything” in production, period. Backup your database maybe? [Intentional off-by-one error] Learn to code a bit too? Extras Brian: What Authors Need to Know About the $1.5 Billion Anthropic Settlement Search LibGen, the Pirated-Books Database That Meta Used to Train AI Simon Willison's list of tools built with the help of LLMs Simon's list of tools that he thinks are genuinely useful and worth highlighting AI Darwin Awards Michael: Python has had async for 10 years -- why isn't it more popular? PyCon Africa Fund Raiser I was on the video stream for about 90 minutes (final 90) Donation page for Python in Africa Jokes: I'm getting the BIOS flavor Is there a seahorse emoji?

Geospatial Data Science Essentials is your hands-on guide to mastering the science of geospatial analytics using Python. Designed for practitioners and enthusiasts alike, this book distills years of experience by wrapping up 101 key concepts from theory to implementation, ensuring you gain a practical understanding of the tools and methods that define the geospatial data science landscape today. Whether you are a seasoned data scientist, a GIS professional, a newcomer to spatial data, or simply a map lover, this book provides you solid foundation to level up your skills. The book is centered around practicalities, as you will explore real-world examples with compact code throughout ten topics and 101 sections. From understanding spatial data structures to leveraging advanced analytical techniques, from spatial networks to machine learning, this book equips you with a wide range of knowledge to navigate and succeed in the rapidly evolving field of geospatial data science.Embrace the journey into geospatial data science with this essential guide and discover the power of Python in unlocking the potential of spatial analytics. Learn more about your ad choices. Visit megaphone.fm/adchoices Support our show by becoming a premium member! https://newbooksnetwork.supportingcast.fm/geography

Geospatial Data Science Essentials is your hands-on guide to mastering the science of geospatial analytics using Python. Designed for practitioners and enthusiasts alike, this book distills years of experience by wrapping up 101 key concepts from theory to implementation, ensuring you gain a practical understanding of the tools and methods that define the geospatial data science landscape today. Whether you are a seasoned data scientist, a GIS professional, a newcomer to spatial data, or simply a map lover, this book provides you solid foundation to level up your skills. The book is centered around practicalities, as you will explore real-world examples with compact code throughout ten topics and 101 sections. From understanding spatial data structures to leveraging advanced analytical techniques, from spatial networks to machine learning, this book equips you with a wide range of knowledge to navigate and succeed in the rapidly evolving field of geospatial data science.Embrace the journey into geospatial data science with this essential guide and discover the power of Python in unlocking the potential of spatial analytics. Learn more about your ad choices. Visit megaphone.fm/adchoices Support our show by becoming a premium member! https://newbooksnetwork.supportingcast.fm/technology

Aubrey Masango host Arthur Mukhuvha, the General Manager of the MTN South Africa Foundation and they talk about his career journey, challenges he faced and more. The Aubrey Masango Show is presented by late night radio broadcaster Aubrey Masango. Aubrey hosts in-depth interviews on controversial political issues and chats to experts offering life advice and guidance in areas of psychology, personal finance and more. All Aubrey’s interviews are podcasted for you to catch-up and listen. Thank you for listening to this podcast from The Aubrey Masango Show. Listen live on weekdays between 20:00 and 24:00 (SA Time) to The Aubrey Masango Show broadcast on 702 https://buff.ly/gk3y0Kj and on CapeTalk between 20:00 and 21:00 (SA Time) https://buff.ly/NnFM3Nk Find out more about the show here https://buff.ly/lzyKCv0 and get all the catch-up podcasts https://buff.ly/rT6znsn Subscribe to the 702 and CapeTalk Daily and Weekly Newsletters https://buff.ly/v5mfet Follow us on social media: 702 on Facebook: https://www.facebook.com/TalkRadio702 702 on TikTok: https://www.tiktok.com/@talkradio702 702 on Instagram: https://www.instagram.com/talkradio702/ 702 on X: https://x.com/Radio702 702 on YouTube: https://www.youtube.com/@radio702 CapeTalk on Facebook: https://www.facebook.com/CapeTalk CapeTalk on TikTok: https://www.tiktok.com/@capetalk CapeTalk on Instagram: https://www.instagram.com/ CapeTalk on X: https://x.com/CapeTalk CapeTalk on YouTube: https://www.youtube.com/@CapeTalk567 See omnystudio.com/listener for privacy information.

In this episode of the Get Plugged In Podcast Series: AI Insights, Dale Hall, Managing Director of the Society of Actuaries Research Institute, is joined by Michael Niemerg, Principal and Director of Data Science and Analytics at Milliman IntelliScript, to explore the urgent and evolving topic of fairness in artificial intelligence, particularly as it applies to insurance underwriting. Michael shares deep insights into the complexities of ensuring fairness in AI-driven models, the implications of generative AI for interpretability, and how actuarial professionals can better align modeling practices with ethical and regulatory standards. The conversation also tackles common misconceptions about AI fairness, the value of additional data in underwriting, and what actuaries need to consider in designing and testing fair models.

Why is AI resorting to blackmail 96% of the time? Today, we're talking to Jon Krohn, host of the Super Data Science podcast and co-founder of YCarrot. We discuss the difference between LLMs and Agentic AI, how businesses can leverage AI for better ROI, and why understanding AI misalignment is crucial for future implementations. All of this right here, right now, on the Modern CTO Podcast!  To learn more about Y Carrot, visit their website here.

Liz Hart is President of Leasing for Newmark's operating businesses in the U.S. and Canada, where she drives the strategy of the firm's leasing platform, leads talent development and recruitment, and helps integrate technology to deliver better outcomes for clients. She also serves on Newmark's Executive Committee, reporting directly to CEO Barry Gosin. With more than 20 years at Newmark, Liz has completed close to 35M square feet of transactions valued at over $4.2 billion. She has consistently ranked among the firm's top producers and was a regular Top Five Producer in Newmark's San Francisco office. Her experience spans advising technology companies from startups to Fortune 50 giants, repositioning large-scale developments that have reshaped skylines, and leading Newmark's Technology & Innovation Practice Group to help landlords and tenants in the TAMI/TMT sectors create spaces that attract and retain talent.(01:16) - State of the Office Market: Shrinking Supply & Turning Point(05:05) - How to Approach Office Leasing in 2025(13:45) - Talent, Culture & Competitive Advantage(15:49) - Data-Driven Leasing & Advisory: Automation vs. Augmentation(18:07) - Feature: CREtech - Join CREtech New York 2025 on Oct 21-22 for the largest Real Estate Meetings program. Qualified Real Estate pros get free full event pass plus up to $800 in travel and hotel costs.(19:39) - Brand Building in Commercial Real Estate(24:32) - Flex Space vs. Traditional Leasing (27:00) - End-to-End Platform: Evolving the Leasing Function(29:02) - In-House vs. Outsourcing Tech & Data(29:41) - Data Sharing & Antitrust: The RealPage Settlement(31:31) - Collaboration Superpower: Steve Jobs

Kevin Werbach interviews DJ Patil, the first U.S. Chief Data Scientist under the Obama Administration, about the evolving role of AI in government, healthcare, and business. Patil reflects on how the mission of government data leadership has grown more critical today: ensuring good data, using it responsibly, and unleashing its power for public benefit. He describes both the promise and the paralysis of today's “big data” era, where dashboards abound, but decision-making often stalls. He highlights the untapped potential of federal datasets, such as the VA's Million Veterans Project, which could accelerate cures for major diseases if unlocked. Yet funding gaps, bureaucratic resistance, and misalignment with Congress continue to stand in the way. Turning to AI, Patil describes a landscape of extraordinary progress: tools that help patients ask the right questions of their physicians, innovations that enhance customer service, and a wave of entrepreneurial energy transforming industries. At the same time, he raises alarms about inequitable access, job disruption, complacency in relying on imperfect systems, and the lack of guardrails to prevent harmful misuse. Rather than relentlessly stepping on the gas in the AI "race," he emphasizes, we need a steering wheel, in the form of public policy, to ensure that AI development serves the public good.  DJ Patil is an entrepreneur, investor, scientist, and public policy leader who served as the first U.S. Chief Data Scientist under the Obama Administration. He has held senior leadership roles at PayPal, eBay, LinkedIn, and Skype, and is currently a General Partner at Greylock Ventures. Patil is recognized as a pioneer in advancing the use of data science to drive innovation, inform policy, and create public benefit. Transcript Ethics of Data Science, Co-Authored by DJ Patil

The consumer goods and retail industries face an overwhelming challenge: too much fragmented data and too little clarity. From mismatched retailer reports to legacy systems that can't keep up with today's SKU volumes, many organizations find themselves bogged down in “data indigestion” instead of actionable insights. Today's guest is Henrique Wakil Moyses, Vice President of Data Science at Crisp. Crisp is a data platform designed for the consumer goods ecosystem, helping brands, retailers, and distributors harmonize fragmented data from multiple sources. By providing real-time visibility into sales, inventory, and supply chain signals, Crisp enables faster, data-driven decisions that reduce waste and improve business outcomes. Henrique joins Emerj Editorial Director Matthew DeMello to break down how CPG and retail leaders can cut through this complexity. He explains why building a data-driven culture is the first barrier to overcome, how to align AI adoption with ROI, and where brands are already seeing the biggest payoffs—such as supply chain optimization, inventory forecasting, and personalized retail experiences. Want to share your AI adoption story with executive peers? Click emerj.com/expert2 for more information and to be a potential future guest on the ‘AI in Business' podcast! This episode is sponsored by Crisp. Learn how brands work with Emerj and other Emerj Media options at emerj.com/ad1.

One of the biggest risks for independent data professionals is spending months or years developing a product or service that nobody wants to buy. The graveyard of failed data science projects is filled with technically brilliant solutions that solved problems no one actually had, leaving their creators with empty bank accounts and bruised egos.In this Value Boost episode, Daniel Bourke joins Dr. Genevieve Hayes to reveal practical strategies for validating data product ideas before investing significant development time, drawing from his experience creating machine learning courses with over 250,000 students and building the Nutrify food education app.This episode uncovers:How to spot genuine market demand before building anything [04:15]The validation strategy that guarantees you win regardless of commercial success [10:16]Why passion projects often create unexpected business opportunities [06:33]The simple approach that turns failed experiments into stepping stones for success [11:50]Guest BioDaniel Bourke is the co-creator of Nutrify, an app described as “Shazam for food”, and teaches machine learning and deep learning at the Zero to Mastery Academy.LinksDaniel's websiteDaniel's YouTube channelConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

Topics covered in this episode: * rathole* * pre-commit: install with uv* A good example of what functools.Placeholder from Python 3.14 allows Converted 160 old blog posts with AI Extras Joke Watch on YouTube About the show Sponsored by DigitalOcean: pythonbytes.fm/digitalocean-gen-ai Use code DO4BYTES and get $200 in free credit 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: rathole A lightweight and high-performance reverse proxy for NAT traversal, written in Rust. An alternative to frp and ngrok. Features High Performance Much higher throughput can be achieved than frp, and more stable when handling a large volume of connections. Low Resource Consumption Consumes much fewer memory than similar tools. See Benchmark. The binary can be as small as ~500KiB to fit the constraints of devices, like embedded devices as routers. On my server, it's currently using about 2.7MB in Docker (wow!) Security Tokens of services are mandatory and service-wise. The server and clients are responsible for their own configs. With the optional Noise Protocol, encryption can be configured at ease. No need to create a self-signed certificate! TLS is also supported. Hot Reload Services can be added or removed dynamically by hot-reloading the configuration file. HTTP API is WIP. Brian #2: pre-commit: install with uv Adam Johnson pre-commit doesn't natively support uv, but you can get around that with pre-commit-uv $ uv tool install pre-commit --with pre-commit-uv Installing pre-commit like this Installs it globally Installs with uv adds an extra plugin “pre-commit-uv” to pre-commit, so that any Python based tool installed via pre-commit also uses uv Very cool. Nice speedup Brian #3: A good example of what functools.Placeholder from Python 3.14 allows Rodrigo Girão Serrão Remove punctuation functionally Also How to use functools.Placeholder, a blog post about it. functools.partial is cool way to create a new function that partially binds some parameters to another function. It doesn't always work for functions that take positional arguments. functools.Placeholder fixes that with the ability to put in placeholders for spots where you want to be able to pass that in from the outer partial binding. And all of this sounds totally obscure without a good example, so thank you to Rodgrigo for coming up with the punctuation removal example (and writeup) Michael #4: Converted 160 old blog posts with AI They were held-hostage at wordpress.com to markdown and integrated them into my Hugo site at mkennedy.codes Here is the chat conversation with Claude Opus/Sonnet. Had to juggle this a bit because the RSS feed only held the last 50. So we had to go back in and web scrape. That resulted in oddies like comments on wordpress that had to be cleaned etc. Whole process took 3-4 hours from idea to “production”duction”. The chat transcript is just the first round getting the RSS → Hugo done. The fixes occurred in other chats. This article is timely and noteworthy: Blogging service TypePad is shutting down and taking all blog content with it This highlights why your domain name needs to be legit, not just tied to the host. I'm looking at you pyfound.blogspot.com. I just redirected blog.michaelckennedy.net to mkennedy.codes Carefully mapping old posts to a new archived area using NGINX config. This is just the HTTP portion, but note the /sitemap.xml and location ~ "^/([0-9]{4})/([0-9]{2})/([0-9]{2})/(.+?)/?$" { portions. The latter maps posts such as https://blog.michaelckennedy.net/2018/01/08/a-bunch-of-online-python-courses/ to https://mkennedy.codes/posts/r/a-bunch-of-online-python-courses/ server { listen 80; server_name blog.michaelckennedy.net; # Redirect sitemap.xml to new domain location = /sitemap.xml { return 301 ; } # Handle blog post redirects for HTTP -> HTTPS with URL transformation # Pattern: /YYYY/MM/DD/post-slug/ -> location ~ "^/([0-9]{4})/([0-9]{2})/([0-9]{2})/(.+?)/?$" { return 301 ; } # Redirect all other HTTP URLs to mkennedy.codes homepage location / { return 301 ; } } Extras Brian: SMS URLs and Draft SMS and iMessage from any computer keyboard from Seth Larson Test and Code Archive is now up, see announcement Michael: Python: The Documentary | An origin story is out! Joke: Do you know him? He is me.

Join Lois Houston and Nikita Abraham as they chat with Yunus Mohammed, a Principal Instructor at Oracle University, about the key stages of AI model development. From gathering and preparing data to selecting, training, and deploying models, learn how each phase impacts AI's real-world effectiveness. The discussion also highlights why monitoring AI performance and addressing evolving challenges are critical for long-term success.   AI for You: https://mylearn.oracle.com/ou/course/ai-for-you/152601/252500   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: 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: Hey everyone! In our last episode, we spoke about generative AI and gen AI agents. Today, we're going to look at the key stages in a typical AI workflow. We'll also discuss how data quality, feedback loops, and business goals influence AI success. With us today is Yunus Mohammed, a Principal Instructor at Oracle University.  01:00 Lois: Hi Yunus! We're excited to have you here! Can you walk us through the various steps in developing and deploying an AI model?  Yunus: The first point is the collect data. We gather relevant data, either historical or real time. Like customer transactions, support tickets, survey feedbacks, or sensor logs. A travel company, for example, can collect past booking data to predict future demand. So, data is the most crucial and the important component for building your AI models. But it's not just the data. You need to prepare the data. In the prepared data process, we clean, organize, and label the data. AI can't learn from messy spreadsheets. We try to make the data more understandable and organized, like removing duplicates, filling missing values in the data with some default values or formatting dates. All these comes under organization of the data and give a label to the data, so that the data becomes more supervised. After preparing the data, I go for selecting the model to train. So now, we pick what type of model fits your goals. It can be a traditional ML model or a deep learning network model, or it can be a generative model. The model is chosen based on the business problems and the data we have. So, we train the model using the prepared data, so it can learn the patterns of the data. Then after the model is trained, I need to evaluate the model. You check how well the model performs. Is it accurate? Is it fair? The metrics of the evaluation will vary based on the goal that you're trying to reach. If your model misclassifies emails as spam and it is doing it very much often, then it is not ready. So I need to train it further. So I need to train it to a level when it identifies the official mail as official mail and spam mail as spam mail accurately.  After evaluating and making sure your model is perfectly fitting, you go for the next step, which is called the deploy model. Once we are happy, we put it into the real world, like into a CRM, or a web application, or an API. So, I can configure that with an API, which is application programming interface, or I add it to a CRM, Customer Relationship Management, or I add it to a web application that I've got. Like for example, a chatbot becomes available on your company's website, and the chatbot might be using a generative AI model. Once I have deployed the model and it is working fine, I need to keep track of this model, how it is working, and need to monitor and improve whenever needed. So I go for a stage, which is called as monitor and improve. So AI isn't set in and forget it. So over time, there are lot of changes that is happening to the data. So we monitor performance and retrain when needed. An e-commerce recommendation model needs updates as there might be trends which are shifting.  So the end user finally sees the results after all the processes. A better product, or a smarter service, or a faster decision-making model, if we do this right. That is, if we process the flow perfectly, they may not even realize AI is behind it to give them the accurate results.  04:59 Nikita: Got it. So, everything in AI begins with data. But what are the different types of data used in AI development?  Yunus: We work with three main types of data: structured, unstructured, and semi-structured. Structured data is like a clean set of tables in Excel or databases, which consists of rows and columns with clear and consistent data information. Unstructured is messy data, like your email or customer calls that records videos or social media posts, so they all comes under unstructured data.  Semi-structured data is things like logs on XML files or JSON files. Not quite neat but not entirely messy either. So they are, they are termed semi-structured. So structured, unstructured, and then you've got the semi-structured. 05:58 Nikita: Ok… and how do the data needs vary for different AI approaches?  Yunus: Machine learning often needs labeled data. Like a bank might feed past transactions labeled as fraud or not fraud to train a fraud detection model. But machine learning also includes unsupervised learning, like clustering customer spending behavior. Here, no labels are needed. In deep learning, it needs a lot of data, usually unstructured, like thousands of loan documents, call recordings, or scan checks. These are fed into the models and the neural networks to detect and complex patterns. Data science focus on insights rather than the predictions. So a data scientist at the bank might use customer relationship management exports and customer demographies to analyze which age group prefers credit cards over the loans. Then we have got generative AI that thrives on diverse, unstructured internet scalable data. Like it is getting data from books, code, images, chat logs. So these models, like ChatGPT, are trained to generate responses or mimic the styles and synthesize content. So generative AI can power a banking virtual assistant trained on chat logs and frequently asked questions to answer customer queries 24/7. 07:35 Lois: What are the challenges when dealing with data?  Yunus: Data isn't just about having enough. We must also think about quality. Is it accurate and relevant? Volume. Do we have enough for the model to learn from? And is my data consisting of any kind of unfairly defined structures, like rejecting more loan applications from a certain zip code, which actually gives you a bias of data? And also the privacy. Are we handling personal data responsibly or not? Especially data which is critical or which is regulated, like the banking sector or health data of the patients. Before building anything smart, we must start smart.  08:23 Lois: So, we've established that collecting the right data is non-negotiable for success. Then comes preparing it, right?  Yunus: This is arguably the most important part of any AI or data science project. Clean data leads to reliable predictions. Imagine you have a column for age, and someone accidentally entered an age of like 999. That's likely a data entry error. Or maybe a few rows have missing ages. So we either fix, remove, or impute such issues. This step ensures our model isn't misled by incorrect values. Dates are often stored in different formats. For instance, a date, can be stored as the month and the day values, or it can be stored in some places as day first and month next. We want to bring everything into a consistent, usable format. This process is called as transformation. The machine learning models can get confused if one feature, like example the income ranges from 10,000 to 100,000, and another, like the number of kids, range from 0 to 5. So we normalize or scale values to bring them to a similar range, say 0 or 1. So we actually put it as yes or no options. So models don't understand words like small, medium, or large. We convert them into numbers using encoding. One simple way is assigning 1, 2, and 3 respectively. And then you have got removing stop words like the punctuations, et cetera, and break the sentence into smaller meaningful units called as tokens. This is actually used for generative AI tasks. In deep learning, especially for Gen AI, image or audio inputs must be of uniform size and format.  10:31 Lois: And does each AI system have a different way of preparing data?  Yunus: For machine learning ML, focus is on cleaning, encoding, and scaling. Deep learning needs resizing and normalization for text and images. Data science, about reshaping, aggregating, and getting it ready for insights. The generative AI needs special preparation like chunking, tokenizing large documents, or compressing images. 11:06 Oracle University's Race to Certification 2025 is your ticket to free training and certification in today's hottest tech. Whether you're starting with Artificial Intelligence, Oracle Cloud Infrastructure, Multicloud, or Oracle Data Platform, this challenge covers it all! Learn more about your chance to win prizes and see your name on the Leaderboard by visiting education.oracle.com/race-to-certification-2025. That's education.oracle.com/race-to-certification-2025. 11:50 Nikita: Welcome back! Yunus, how does a user choose the right model to solve their business problem?  Yunus: Just like a business uses different dashboards for marketing versus finance, in AI, we use different model types, depending on what we are trying to solve. Like classification is choosing a category. Real-world example can be whether the email is a spam or not. Use in fraud detection, medical diagnosis, et cetera. So what you do is you classify that particular data and then accurately access that classification of data. Regression, which is used for predicting a number, like, what will be the price of a house next month? Or it can be a useful in common forecasting sales demands or on the cost. Clustering, things without labels. So real-world examples can be segmenting customers based on behavior for targeted marketing. It helps discovering hidden patterns in large data sets.  Generation, that is creating new content. So AI writing product description or generating images can be a real-world example for this. And it can be used in a concept of generative AI models like ChatGPT or Dall-E, which operates on the generative AI principles. 13:16 Nikita: And how do you train a model? Yunus: We feed it with data in small chunks or batches and then compare its guesses to the correct values, adjusting its thinking like weights to improve next time, and the cycle repeats until the model gets good at making predictions. So if you're building a fraud detection system, ML may be enough. If you want to analyze medical images, you will need deep learning. If you're building a chatbot, go for a generative model like the LLM. And for all of these use cases, you need to select and train the applicable models as and when appropriate. 14:04 Lois: OK, now that the model's been trained, what else needs to happen before it can be deployed? Yunus: Evaluate the model, assess a model's accuracy, reliability, and real-world usefulness before it's put to work. That is, how often is the model right? Does it consistently perform well? Is it practical in the real world to use this model or not? Because if I have bad predictions, doesn't just look bad, it can lead to costly business mistakes. Think of recommending the wrong product to a customer or misidentifying a financial risk.  So what we do here is we start with splitting the data into two parts. So we train the data by training data. And this is like teaching the model. And then we have got the testing data. This is actually used for checking how well the model has learned. So once trained, the model makes predictions. We compare the predictions to the actual answers, just like checking your answer after a quiz. We try to go in for tailored evaluation based on AI types. Like machine learning, we care about accuracy in prediction. Deep learning is about fitting complex data like voice or images, where the model repeatedly sees examples and tunes itself to reduce errors. Data science, we look for patterns and insights, such as which features will matter. In generative AI, we judge by output quality. Is it coherent, useful, and is it natural?  The model improves with the accuracy and the number of epochs the training has been done on.  15:59 Nikita: So, after all that, we finally come to deploying the model… Yunus: Deploying a model means we are integrating it into our actual business system. So it can start making decisions, automating tasks, or supporting customer experiences in real time. Think of it like this. Training is teaching the model. Evaluating is testing it. And deployment is giving it a job.  The model needs a home either in the cloud or inside your company's own servers. Think of it like putting the AI in place where it can be reached by other tools. Exposed via API or embedded in an app, or you can say application, this is how the AI becomes usable.  Then, we have got the concept of receives live data and returns predictions. So receives live data and returns prediction is when the model listens to real-time inputs like a user typing, or user trying to search or click or making a transaction, and then instantly, your AI responds with a recommendation, decisions, or results. Deploying the model isn't the end of the story. It is just the beginning of the AI's real-world journey. Models may work well on day one, but things change. Customer behavior might shift. New products get introduced in the market. Economic conditions might evolve, like the era of COVID, where the demand shifted and the economical conditions actually changed. 17:48 Lois: Then it's about monitoring and improving the model to keep things reliable over time. Yunus: The monitor and improve loop is a continuous process that ensures an AI model remains accurate, fair, and effective after deployment. The live predictions, the model is running in real time, making decisions or recommendations. The monitor performance are those predictions still accurate and helpful. Is latency acceptable? This is where we track metrics, user feedbacks, and operational impact. Then, we go for detect issues, like accuracy is declining, are responses feeling biased, are customers dropping off due to long response times? And the next step will be to reframe or update the model. So we add fresh data, tweak the logic, or even use better architectures to deploy the uploaded model, and the new version replaces the old one and the cycle continues again. 18:58 Lois: And are there challenges during this step? Yunus: The common issues, which are related to monitor and improve consist of model drift, bias, and latency of failures. In model drift, the model becomes less accurate as the environment changes. Or bias, the model may favor or penalize certain groups unfairly. Latency or failures, if the model is too slow or fails unpredictably, it disrupts the user experience. Let's take the loan approvals. In loan approvals, if we notice an unusually high rejection rate due to model bias, we might retrain the model with more diverse or balanced data. For a chatbot, we watch for customer satisfaction, which might arise due to model failure and fine-tune the responses for the model. So in forecasting demand, if the predictions no longer match real trends, say post-pandemic, due to the model drift, we update the model with fresh data.  20:11 Nikita: Thanks for that, Yunus. Any final thoughts before we let you go? Yunus: No matter how advanced your model is, its effectiveness depends on the quality of the data you feed it. That means, the data needs to be clean, structured, and relevant. It should map itself to the problem you're solving. If the foundation is weak, the results will be also. So data preparation is not just a technical step, it is a business critical stage. Once deployed, AI systems must be monitored continuously, and you need to watch for drops in performance for any bias being generated or outdated logic, and improve the model with new data or refinements. That's what makes AI reliable, ethical, and sustainable in the long run. 21:09 Nikita: Yunus, thank you for this really insightful session. If you're interested in learning more about the topics we discussed today, go to mylearn.oracle.com and search for the AI for You course.  Lois: That's right. You'll find skill checks to help you assess your understanding of these concepts. In our next episode, we'll discuss the idea of buy versus build in the context of AI. Until then, this is Lois Houston… Nikita: And Nikita Abraham, signing off! 21:39 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.

Adatépítész -az első magyar datapodcast Minden ami hír, érdekesség, esemény vagy tudásmorzsa az  adat, datascience, adatbányászat és hasonló kockaságok világából. Become a Patron! UPDATE: az Adás felvétele óta eltelt időben sem kaptam a cikk szerzőitől semmiféle választ az érdeklődésemre.

Twenty years after a scrappy newsroom team hacked together a framework to ship stories fast, Django remains the Python web framework that ships real apps, responsibly. In this anniversary roundtable with its creators and long-time stewards: Simon Willison, Adrian Holovaty, Will Vincent, Jeff Triplet, and Thibaud Colas, we trace the path from the Lawrence Journal-World to 1.0, DjangoCon, and the DSF; unpack how a BSD license and a culture of docs, tests, and mentorship grew a global community; and revisit lessons from deployments like Instagram. We talk modern Django too: ASGI and async, HTMX-friendly patterns, building APIs with DRF and Django Ninja, and how Django pairs with React and serverless without losing its batteries-included soul. You'll hear about Django Girls, Djangonauts, and the Django Fellowship that keep momentum going, plus where Django fits in today's AI stacks. Finally, we look ahead at the next decade of speed, security, and sustainability. Episode sponsors Talk Python Courses Python in Production Links from the show Guests Simon Willison: simonwillison.net Adrian Holovaty: holovaty.com Will Vincent: wsvincent.com Jeff Triplet: jefftriplett.com Thibaud Colas: thib.me Show Links Django's 20th Birthday Reflections (Simon Willison): simonwillison.net Happy 20th Birthday, Django! (Django Weblog): djangoproject.com Django 2024 Annual Impact Report: djangoproject.com Welcome Our New Fellow: Jacob Tyler Walls: djangoproject.com Soundslice Music Learning Platform: soundslice.com Djangonaut Space Mentorship for Django Contributors: djangonaut.space Wagtail CMS for Django: wagtail.org Django REST Framework: django-rest-framework.org Django Ninja API Framework for Django: django-ninja.dev Lawrence Journal-World: ljworld.com Watch this episode on YouTube: youtube.com Episode #518 deep-dive: talkpython.fm/518 Episode transcripts: talkpython.fm Developer Rap Theme Song: Served in a Flask: talkpython.fm/flasksong --- Stay in touch with us --- Subscribe to Talk Python on YouTube: youtube.com Talk Python on Bluesky: @talkpython.fm at bsky.app Talk Python on Mastodon: talkpython Michael on Bluesky: @mkennedy.codes at bsky.app Michael on Mastodon: mkennedy

This month, we explore how data science and AI are transforming the wine industry—from vineyard planting and grape harvesting to customer engagement. Can advanced technologies help winemakers enhance quality, promote sustainability, and better match wines to consumers—all while preserving the essential human touch? Might these innovations be applied to other products as well? Join us as we discuss these questions and more with industry leaders Kia Behnia, CEO and co-founder of Scout, and Katerina Axelsson, CEO and founder of Tastry. Pour yourself a glass and tune in as we uncork the intersection of data, AI, and the art of winemaking. Our Guests: Kia Behnia is CEO and co-founder of Scout, an AI-powered analytics platform built for precision viticulture, and proprietor of Kiatra Vineyards and Neotempo Wines.  Katerina Axelsson is CEO and founder of Tastry, a sensory-sciences company that blends advanced analytical chemistry, machine learning, and AI to predict consumer preferences—especially in wine.

A nostalgic dive into the rise and fall of true hacker culture - from MIT's curious tinkerers to today's hustle-obsessed "founders." Plus, why IRC was peak internet and what we lost when convenience killed community. For anyone who misses when coding was about elegance, not exits.RetryClaude can make mistakes. Please double-check responses. Interesting link https://www.twitch.tv/tsoding/about Sponsors DSH is proudly sponsored by Amethix Technologies. At the intersection of ethics and engineering, Amethix creates AI systems that don't just function—they adapt, learn, and serve. With a focus on dual-use innovation, Amethix is shaping a future where intelligent machines extend human capability, not replace it. Discover more at amethix.com   DSH is brought to you by Intrepid AI. From drones to satellites, Intrepid AI gives engineers and defense innovators the tools to prototype, simulate, and deploy autonomous systems with confidence. Whether it's in the sky, on the ground, or in orbit—if it's intelligent and mobile, Intrepid helps you build it. Learn more at intrepid.ai     ✨ Connect with us!

A statistician walks into a bar, and a comedy and art show begins. Creative work for scholars can extend beyond novel research and application. In today's episode of stats and stories, we see how the intersection between interest in statistics and art, as well as the intersection of statistics and comedy, with Dr Greg Matthews. Dr. Matthews is Associate Professor of Statistics and Director of the Center for Data Science and Consulting at Loyola University. He also is a data artist who developed and promoted the Data Art Show, which debuted at the 2016 Joint Statistical Meetings. He performs with the Uncontrolled Variables comedy troupe at the Lincoln Lodge in Chicago and you can see his data art, links to his comedy performance, and much more at his website, Stats in the Wild.

AI in software development sounds like a dream, faster coding, cleaner refactoring, and technical reports that actually make sense to stakeholders. But, what's the bad news in the classic good news/bad news scenario? Poisoned training data, compliance risks, and systems that are brittle and will not scale.  This week on Feds At The Edge, Alex Gromadzki, Assistant Director of Data Science at US GAO, and Steven Toy, Senior Director, Cloud Infrastructure for ICF, unpack the opportunities and pitfalls of generative AI in federal software development. From source-citing AI to data security in the software lifecycle, they reveal why small, testable use cases may be the smartest way forward.   Listen now on your favorite podcast platform to hear how federal leaders can balance innovation with responsibility as AI reshapes the software development life cycle.          

Rishard Rameez is the Co‑Founder and CEO of Zown, an AI‑powered real estate platform that makes homeownership more accessible and affordable. Zown was born from a viral Reddit post where Rishard shared his frustration over paying over $70K in real estate commissions. The outpouring of support inspired him to flip the model: instead of paying big commissions, Zown gives buyers significant upfront cash to help with their down payment and closing costs, while offering sellers flat fees. This customer‑first model has driven rapid growth, with Zown processing over $300 million in transactions and becoming Canada's fastest‑growing real estate brokerage. The platform has recently launched in California and continues expanding across North America. Rishard sparked a movement by transforming personal pain into an industry‑changing solution.(02:17) - The Broken Home Buying Process(03:02) - It All Started with a Viral Reddit Post (05:39) - Early Pivot from Flat Fee Model(14:59) - Unbundling Real Estate Services(18:06) - Feature: Blueprint - The Future of Real Estate - Register for 2025: The Premier Event for Industry Executives, Real Estate & Construction Tech Startups and VC's, at The Venetian, Las Vegas on Sep. 16th-18th, 2025.(19:00) - Feature: Meow - Business banking, with interest: Unlock a high-yield business checking account that pays up to 3.52%.(20:31) - Zone's Growth Journey(28:47) - Customer Acquisition Strategy(30:36) - Recent Seed Round(32:42) - Why Own vs. Rent a Home(34:52) - Collaboration Superpower: Muhammad and Jesus Christ

Celebrate a decade of innovation, learning, and connection in the Alteryx Community! In this special 10th anniversary episode of Alter Everything, we hear from you as we explore the stories and milestones that have defined the Alteryx Community over the past ten years. Hear firsthand accounts from users like you (maybe even you reading this) whose lives and careers have been transformed through mentorship, career advancement, or lifelong friendships. This episode highlights the power of Community in the world of data analytics. Join us as we honor the people, stories, and achievements that make the Alteryx Community truly special.Guests: Matt Rotundo, Engagement Engineer @Alteryx - @AlteryxMatt, LinkedInAlex Gross, Sr. Process Analyst @ Siemens - @grossal, LinkedInNicole Johnson, Sr. Manager Product Management @ Alteryx - @NicoleJ, LinkedInMatt Montgomery, Data Sherpa @ Montgomery Solutions - @mmontgomery, LinkedInCalvin Tang, Group Manager, Business Solutions & Enablement @ Prudential PLC - @Caltang, LinkedInSamantha Clifton, Sr. Sales Engineer @ Alteryx - @Samantha_Jayne, LinkedInLuke Cornetta, Sr. Director @ Alvarez and Marsal - @LukeC, LinkedInBen Stringer, Data Consultant @ Bulien - @BS_THE_ANALYST, LinkedInRoan Pilsworth, Data Consultant @ Bulien - @pilsner, LinkedInAlex Abi-Najm, Solutions and Enablement Lead @ Aimpoint Digital - @alexnajm, LinkedInShan Miralles, Quantitative Analyst @ JP Morgan Chase - @shancmiralles, LinkedInDan Menke, Community Ops Sr. Manager @ Alteryx - @DanM, LinkedInMegan Bowers, Sr. Content Manager @ Alteryx - @MeganBowers, LinkedInShow notes: Inspire ConferenceAlteryx ACE ProgramAdvent of CodeWeekly Challenges and Cloud QuestsAlteryx User GroupsAlteryx AcademyAlteryx Interactive LessonsAlteryx CertificationSparkEdRoad to Inspire Interested in sharing your feedback with the Alter Everything team? Take our feedback survey here!This episode was produced by Megan Bowers, Mike Cusic, and Matt Rotundo. Special thanks to Andy Uttley for the theme music.

The traditional career path of climbing the corporate ladder no longer appeals to many data scientists - who crave freedom and ownership of their work. Yet the leap from employment to independence can feel risky and uncertain, especially without a clear roadmap for success.In this episode, Daniel Bourke joins Dr. Genevieve Hayes to share his journey from machine learning engineer to successful independent data professional before age 30, revealing the practical steps and mindset shifts needed to transform technical skills into sustainable freedom.In this episode, you'll discover:Why embracing the "permissionless economy" is crucial for independent success [14:59]The power of "starting the job before you have it" [12:17]Why building your own website is the foundation for long-term independent success [24:35]A practical approach to opportunity selection that accelerates career momentum [17:27]Guest BioDaniel Bourke is the co-creator of Nutrify, an app described as “Shazam for food”, and teaches machine learning and deep learning at the Zero to Mastery Academy.LinksDaniel's websiteDaniel's YouTube channelConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

Send us a textAI-Driven Business Transformation with Hitachi SolutionsThe Evolving Role of Data Science in AI: Hosted by Laurel Greszler, she visits with Hitachi Solutions Data Science & AI Resident Experts Fred Heller and Mike Kirkpatrick to discuss how data science is foundational to unlocking business value from vast, underutilized data assets. They emphasize that data scientists transform raw data into actionable insights, enabling organizations to automate and optimize decision-making processes. AI as a Strategic Business Enabler: The team highlights that AI, particularly generative AI (GenAI), is not a magic solution but a powerful tool that, when applied thoughtfully, accelerates business outcomes. They stress the importance of setting realistic expectations and integrating AI into existing analytical best practices. Three Core AI Solution Categories:Custom AI Chatbots: Tailored chatbots using Retrieval Augmented Generation (RAG) architectures allow dynamic interrogation of large text datasets, providing users with precise, context-aware answers. This approach leverages Azure cloud services for security and scalability, aligning with Microsoft investments. Document Processing Automation: AI-driven document processing dramatically reduces manual effort, cycle times, and costs by automating the extraction and summarization of information from large volumes of documents. This is especially impactful in industries burdened by compliance and documentation requirements. AI-Powered Document Generation: Generative AI can rapidly produce high-quality first drafts of complex documents (e.g., proposals, reports), giving teams a significant head start and freeing up time for higher-value work. This solution is particularly valuable for organizations that produce lengthy, formulaic documents. Custom vs. Out-of-the-Box Solutions: The conversation distinguishes between out-of-the-box tools like Microsoft Copilot and custom AI builds. While standard solutions offer quick wins, custom builds are essential for complex, industry-specific needs, offering better long-term value and integration with existing business processes. Industry-Agnostic Impact: The team shares that these AI solutions are applicable across industries—from manufacturing to pharmaceuticals—wherever there is a need to process or generate large volumes of information. The common thread is the desire to reduce manual reading and writing, improve efficiency, and empower employees to focus on higher-value tasks. Customer-Centric Approach: Hitachi Solutions' methodology starts with understanding the customer's true business challenge, ensuring that AI solutions are tailored to deliver measurable impact. Advisory workshops and collaborative design sessions are used to align technology with business goals. Microsoft Azure Expertise: As a Microsoft-dedicated partner, Hitachi Solutions leverages Azure's robust AI and data services to deliver secure, scalable, and future-ready solutions, ensuring customers maximize their existing technology investments. Key Takeaway: AI and data science, when strategically applied, can transform business operations, reduce costs, and enhance employee satisfaction. Hitachi Solutions offers deep expertise in custom AI development, ensuring solutions are both innovative and aligned with each customer's unique needs. global.hitachi-solutions.com

Topics covered in this episode: * pypistats.org was down, is now back, and there's a CLI* * State of Python 2025* * wrapt: A Python module for decorators, wrappers and monkey patching.* pysentry 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. Brian #1: pypistats.org was down, is now back, and there's a CLI pypistats.org is a cool site to check the download stats for Python packages. It was down for a while, like 3 weeks? A couple days ago, Hugo van Kemenade announced that it was back up. With some changes in stewardship “pypistats.org is back online!

We're bringing back one of our most downloaded episodes ever – a deep dive into how adverse events should be analyzed properly. This conversation with Jan Beyersmann and Kaspar Rufibach is packed with methodological insights and practical implications for statisticians working in clinical trials. Adverse event (AE) analysis has long been approached differently from efficacy analysis, often using overly simplistic methods that can bias results. In this episode, we discuss why that's a problem – and how the SAVVY collaboration (Survival analysis for AdVerse events with Varying follow-up times) is pushing the field forward. Together with academia and multiple pharma companies, this collaboration tackled the issue of AE analysis using real randomized trial data, not just simulations. The findings show how common methods can underestimate or overestimate event probabilities and how established statistical methods can be applied more consistently to ensure fair benefit–risk assessments. If you've ever wondered whether your approach to safety analysis is leading to misleading conclusions, this episode is a must-listen.

Agentic AI programming is what happens when coding assistants stop acting like autocomplete and start collaborating on real work. In this episode, we cut through the hype and incentives to define “agentic,” then get hands-on with how tools like Cursor, Claude Code, and LangChain actually behave inside an established codebase. Our guest, Matt Makai, now VP of Developer Relations at DigitalOcean, creator of Full Stack Python and Plushcap, shares hard-won tactics. We unpack what breaks, from brittle “generate a bunch of tests” requests to agents amplifying technical debt and uneven design patterns. Plus, we also discuss a sane git workflow for AI-sized diffs. You'll hear practical Claude tips, why developers write more bugs when typing less, and where open source agents are headed. Hint: The destination is humans as editors of systems, not just typists of code. Episode sponsors Posit Talk Python Courses Links from the show Matt Makai: linkedin.com Plushcap Developer Content Analytics: plushcap.com DigitalOcean Gradient AI Platform: digitalocean.com DigitalOcean YouTube Channel: youtube.com Why Generative AI Coding Tools and Agents Do Not Work for Me: blog.miguelgrinberg.com AI Changes Everything: lucumr.pocoo.org Claude Code - 47 Pro Tips in 9 Minutes: youtube.com Cursor AI Code Editor: cursor.com JetBrains Junie: jetbrains.com Claude Code by Anthropic: anthropic.com Full Stack Python: fullstackpython.com Watch this episode on YouTube: youtube.com Episode #517 deep-dive: talkpython.fm/517 Episode transcripts: talkpython.fm Developer Rap Theme Song: Served in a Flask: talkpython.fm/flasksong --- Stay in touch with us --- Subscribe to Talk Python on YouTube: youtube.com Talk Python on Bluesky: @talkpython.fm at bsky.app Talk Python on Mastodon: talkpython Michael on Bluesky: @mkennedy.codes at bsky.app Michael on Mastodon: mkennedy

Python's data stack is getting a serious GPU turbo boost. In this episode, Ben Zaitlen from NVIDIA joins us to unpack RAPIDS, the open source toolkit that lets pandas, scikit-learn, Spark, Polars, and even NetworkX execute on GPUs. We trace the project's origin and why NVIDIA built it in the open, then dig into the pieces that matter in practice: cuDF for DataFrames, cuML for ML, cuGraph for graphs, cuXfilter for dashboards, and friends like cuSpatial and cuSignal. We talk real speedups, how the pandas accelerator works without a rewrite, and what becomes possible when jobs that used to take hours finish in minutes. You'll hear strategies for datasets bigger than GPU memory, scaling out with Dask or Ray, Spark acceleration, and the growing role of vector search with cuVS for AI workloads. If you know the CPU tools, this is your on-ramp to the same APIs at GPU speed. Episode sponsors Posit Talk Python Courses Links from the show RAPIDS: github.com/rapidsai Example notebooks showing drop-in accelerators: github.com Benjamin Zaitlen - LinkedIn: linkedin.com RAPIDS Deployment Guide (Stable): docs.rapids.ai RAPIDS cuDF API Docs (Stable): docs.rapids.ai Asianometry YouTube Video: youtube.com cuDF pandas Accelerator (Stable): docs.rapids.ai Watch this episode on YouTube: youtube.com Episode #516 deep-dive: talkpython.fm/516 Episode transcripts: talkpython.fm Developer Rap Theme Song: Served in a Flask: talkpython.fm/flasksong --- Stay in touch with us --- Subscribe to Talk Python on YouTube: youtube.com Talk Python on Bluesky: @talkpython.fm at bsky.app Talk Python on Mastodon: talkpython Michael on Bluesky: @mkennedy.codes at bsky.app Michael on Mastodon: mkennedy

Topics covered in this episode: pyx - optimized backend for uv * Litestar is worth a look* * Django remake migrations* * django-chronos* Extras Joke Watch on YouTube About the show Python Bytes 445 Sponsored by Sentry: pythonbytes.fm/sentry - Python Error and Performance Monitoring 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: pyx - optimized backend for uv via John Hagen (thanks again) I'll be interviewing Charlie in 9 days on Talk Python → Sign up (get notified) of the livestream here. Not a PyPI replacement, more of a middleware layer to make it better, faster, stronger. pyx is a paid service, with maybe a free option eventually. Brian #2: Litestar is worth a look James Bennett Michael brought up Litestar in episode 444 when talking about rewriting TalkPython in Quart James brings up scaling - Litestar is easy to split an app into multiple files Not using pydantic - You can use pydantic with Litestar, but you don't have to. Maybe attrs is right for you instead. Michael brought up Litestar seems like a “more batteries included” option. Somewhere between FastAPI and Django. Brian #3: Django remake migrations Suggested by Bruno Alla on BlueSky In response to a migrations topic last week django-remake-migrations is a tool to help you with migrations and the docs do a great job of describing the problem way better than I did last week “The built-in squashmigrations command is great, but it only work on a single app at a time, which means that you need to run it for each app in your project. On a project with enough cross-apps dependencies, it can be tricky to run.” “This command aims at solving this problem, by recreating all the migration files in the whole project, from scratch, and mark them as applied by using the replaces attribute.” Also of note The package was created with Copier Michael brought up Copier in 2021 in episode 219 It has a nice comparison table with CookieCutter and Yoeman One difference from CookieCutter is yml vs json. I'm actually not a huge fan of handwriting either. But I guess I'd rather hand write yml. So I'm thinking of trying Copier with my future project template needs. Michael #4: django-chronos Django middleware that shows you how fast your pages load, right in your browser. Displays request timing and query counts for your views and middleware. Times middleware, view, and total per request (CPU and DB). Extras Brian: Test & Code 238: So Long, and Thanks for All the Fish after 10 years, this is the goodbye episode Michael: Auto-activate Python virtual environment for any project with a venv directory in your shell (macOS/Linux): See gist. Python 3.13.6 is out. Open weight OpenAI models Just Enough Python for Data Scientists Course The State of Python 2025 article by Michael Joke: python is better than java

Bob Rudis, VP Data Science from GreyNoise, is sharing some insights into their work on "Early Warning Signals: When Attacker Behavior Precedes New Vulnerabilities." New research reveals a striking trend: in 80% of cases, spikes in malicious activity against enterprise edge technologies like VPNs and firewalls occurred weeks before related CVEs were disclosed. The report breaks down this “6-week critical window,” highlighting which vendors show the strongest early-warning patterns and offering tactical steps defenders can take when suspicious spikes emerge. These findings reveal how early attacker activity can be transformed into actionable intelligence, enabling defenders to anticipate and neutralize threats before vulnerabilities are publicly disclosed. Complete our annual ⁠⁠⁠audience survey⁠⁠⁠ before August 31. The research can be found here: Early Warning Signals: When Attacker Behavior Precedes New Vulnerabilities Learn more about your ad choices. Visit megaphone.fm/adchoices

A young computer scientist and two colleagues show that searches within data structures called hash tables can be much faster than previously deemed possible. The story How Undergraduate Upends a 40-Year-Old Data Science Conjecture first appeared on Quanta Magazine.