The Voice Box

What does a person do after 35 years in tech, including 12 years at Amazon, serving as VP of Kindle, Echo, and Alexa, and Jeff Bezos's TA (“shadow”)? Well, in Ian Freed's case, you launch a voice-oriented company that will make a real difference. Bamboo Learning is the leader in conversation-based learning for children.

Yactraq bills itself as the value leader in conversational AI / speech analytics for contact centers. Jeff and Darin talk with Jeh Daruvula, Benjamin Land, Josh Ayres, and MK about how speech technology is being used in contact centers to improve the customer experience.

Speechly provides accurate, real-time speech recognition and natural language understanding tools under one flexible API.  In this episode, we speak with founder Otto Söderlund about what that means for consumers and businesses. And we talk a little bit about food & speech. Here is a demo of the kind of food ordering by voice that Otto describes in the podcast.

How are mom's carrot cookies similar to the Eastern Min dialect of Chinese? Daniel Zheng is on a mission to preserve Eastern Min with the help of speech technology. Despite having about 10 million speakers, the language is in danger of extinction in a generation or two, and Daniel is on a mission to reverse that. He founded ZFC to help with education and technology in support of Eastern Min preservation. And before you ask, here's the recipe for... Mom's Carrot Cookies Ingredients: 3/4 C (170 g) shortening3/4 C (150 g) sugar1 egg1 teaspoon (5 ml) vanilla extract1 C (240 ml) cooked, mashed, cooled carrots2 C (240 g) white flour2 teaspoons (10ml) salt1/2 C (120 ml) chopped nuts (walnuts or pecans) Icing: 1 box (450 g) powdered sugarThe rind of 1 orange, zestedThe juice of 1 orange Directions: Heat oven to 350°F (175 C).Cream together sugar and shortening.Add egg, vanilla, and carrots & mix.Add dry ingredients & mix.Bake at 350°F (175 C) for 10 to 12 minutes.Frost with icing Let us know if you try them!

Ahmed Bouzid has been building and delivering voice solutions his entire career. His company, Witlingo, helps other companies engage their audiences with voice on several levels. That might involve helping companies develop voice skills & interfaces, or providing tools to allow customers to share voice reviews and testimonials.

Constant Companion uses empathy, data and AI powered voice and video solutions to reinvent the way we care for the elderly and those who need in-home care. Their systems keep people more connected, engaged and protected, and help people live safer and more independent lives with dignity. Founder Mark Gray talks with us about the role voice plays in this process.

Rene Arvin is the co-founder of VoiceScript, a new company that is transforming - and saving - the court transcription industry. Rene talks with Darin and Jeff about the origins of VoiceScript, and what we might expect to see in courtrooms and legal deposition proceedings of the near future.

As a platform enabler, Ken Sutton of Yobe see a future where voice is moving from a "nice to have" user interface to a "need to have" revenue generator. Yobe's AI-powered platform is purpose-built for live crowds and noisy environments to identify and decode human voices. Modeled on human hearing, Yobe's signal processing techniques substantially increase SNRs (signal-to-noise ratio) in noisy environments allowing the ability to decipher emotion, intent, mood, and other biological markers for an added layer of meaning. In this episode, Ken talks with Darin & Jeff about the challenges of understanding speech in real-world environments.

Igor Jablokov is a force of nature. He previously founded and ran Yap, where our host Jeff Adams ran the research group. Now he is the founder and CEO of Pryon, an artificial intelligence (AI) company delivering an enterprise knowledge management platform. Pryon's natural language processing (NLP) engine ingests and transforms data into experiences that solve critical business challenges. In this episode, he talks with Darin and Jeff about Pryon's enterprise AI technology, and about tech in general.

Roberto Sicconi discusses the use of voice (and other) tech that Dreyev is bringing to market to help drivers stay safe. Dreyev offers an in-vehicle “digital copilot” device that evaluates drivers for risky behavior. Through Computer Vision and Machine Learning, the Dreyev system analyzes driver conditions such as head pose and eyelid closure to detect distracted or drowsy driving and issues real-time voice alerts in the case of dangerous conditions.

Babbly is building an AI powered platform that will enable parents to monitor their child's development in early years of life. Babbly can analyze a child's speech and language skills and empower parents to track their baby's development like they would with sleep and body temperature. Maryam Nabavi is a mother, and a strategic Leader with 10+ yrs of experience in product innovation & consumer electronics. Carla Margalef Bentabol has deep experience in software development, artificial intelligence, and natural language processing. On this episode of The Voice Box, Maryam and Carla describe their vision to Darin and Jeff.

Rev.com is a leading provider of transcription services, with a "gig economy" approach to transcription. Miguel Jetté has been leading Rev's speech technology team for 5 years and has fascinating insights into the melding of human and computer resources to increase accuracy and productivity of speech transcription.

On this episode, we talk with Xavier Anguerra about the English language accent coach, ELSA. Xavier has a reputation in the speech industry to be reckoned with. He has over 20 years of experience in speech and signal processing technologies, has published over 100 peer-reviewed research papers, and holds multiple patents in the areas of multimedia, machine learning, deep learning and artificial intelligence.

Tune in to hear the story of how a missing persons report led to the development of local voice control for hospitals and homes. SmartOR is giving voice to a completely end to end smart home. From lighting to streaming services, they are giving users control without sacrificing privacy. In the hospital OR space, they are using voice to assist with arthroscopic surgery. In the OR of the future. The surgeon will be able to control everything by voice. Will the voice controls of the future always require a web connection, and the transfer of your speech and data to the cloud?

Ben Walker founded Transcription Outsourcing in 2010 as a medical transcription company, and they have now diversified into legal, law enforcement, academic, insurance, and many other industries. Unlike other guests on this podcast, they do things the old-fashioned way -- with human transcribers. In this episode, we explore some of the cases where the human touch is necessary, and when it might make sense to apply a hybrid human+computer approach to transcription, and we discuss some of the challenges computers will need to overcome to reach parity.

This episode, we are fortunate to have Beth Porter as our guest. Beth leads the team at Riff Analytics, where they model the conversations of small groups working collaboratively in order to give participants real-time feedback and insights about interactions over time. Have you ever been in a Zoom meeting (or any meeting) that went off the rails, or that was dominated by one person? Riff is helping to fix that problem through their tools that use AI to measure and augment human interactions, helping people and organizations communicate, collaborate, and innovate better.

Don Wright joins us this week to talk about Clarigent, and their new product, Clarity. Clarity gives clinicians an important new tool, based on AI technology, to track suicide risk and mental health outcomes. Don has spent over 30 years in the healthcare and informatics space. Prior to Clarigent Health, he led Assurex Health, which grew to over 500 employees and was acquired in 2016 by Myriad Genetics for $225 million. When Don left Assurex after the transition period, he wanted to continue working in the mental health field, and to build another company that helped people. Lack of funding and other factors including stigma have caused mental health to lag behind other areas of medical practice when it comes to integrating technology and science. In 2018, he established Clarigent Health with the goal of bringing science to mental health. The core technology builds on foundational research conducted at Cincinnati Children's Hospital Medical Center led by Dr. John Pestian and Dr. Tracy Glauser. Don's co-founder, Bill Haynes had previously worked with Don, John and Tracy during the Assurex days, including Assurex sponsoring some of the research that now constitutes the core technology at Clarigent. Although Don has worked in and advocated for mental healthcare for a long time, three years ago this purpose became more deeply personal. His son Justin died by suicide just as Clarigent was developing the commercial version of the suicide prevention technology. The memory of Justin's life has made this suicide prevention work a more personal mission for Don and the entire team. Darin and Jeff are honored to have Don on The Voice Box, and Cobalt Speech is happy to support Clarigent with speech technology.

Shyamala Paraga heads speech interfaces for Ford Motors. She joins us this week to talk about how we'll converse with our cars in the future. Shyamala has 20 years of experience in UX design, 9 years of which have been focused on voice interaction and VUI. Shyamala is a leading figure in the field of voice interfaces. See her web site here, including links to her books and publications. Apologies to both Shyamala and our listeners: This was the first interview we recorded for The Voice Box, and we had some troubles with the audio. This is a great listen, with important content, but we admit the sound quality is less than perfect.

Bruce Rasa wants to revolutionize information capture and flow within the world of agriculture & food. His company, AgVoice is simplifying the capture of data in real-time at the source completely hands-free. They are trusted experts in voice-to-data collection across a global network of partners who are among the largest seed, dairy, speciality and row crop organizations in the world. Listen in on this episode as Jeff & Darin talk to Bruce about his vision for a future where consumers will have greater clarity, confidence, and transparency about the origin and processing of the food they eat.

Check out the future of interactive storytelling with the award winning interactive writer and founder / CEO of EarReality, Christian Mahnke. Ear-Reality is a market leader in interactive audio stories for voice. As an agency, they offer services for companies and brands. As a publisher they create and publish our own interactive stories. Ear-Reality's TWIST (The Wonderful Interactive Storytelling Tool) allows writers and content creators to create and publish interactive audio stories easily for Amazon Alexa, Google Assistant and Samsung Bixby.

Daniel Whitenack (aka Data Dan) is a Ph.D. trained data scientist working with SIL International on NLP and speech technology for local languages in emerging markets. He has more than ten years of experience developing and deploying machine learning models at scale. Daniel co-hosts the Practical AI podcast, has spoken at conferences around the world (Applied Machine Learning Days, O'Reilly AI, QCon AI, GopherCon, KubeCon, and more), and occasionally teaches data science/analytics at Purdue University. Dan talks with Jeff and Darin about the challenges of bringing language technology and resources to all the people of the world. Also check out the Practical AI podcast, which Dan co-hosts.

Henry O'Connell, CEO and co-founder of Canary Speech, joins us on The Voice Box to discuss their efforts to monitor our health and condition through our speech. This is a particularly interesting episode for our co-host Jeff Adams, who is also a co-founder at Canary Speech. Henry and Jeff have been friends for 35 years. Interesting Links: Check out this recent article highlighting Canary speech, and naming it among the 15 most promising startups. --- The Voice Box is sponsored by Cobalt Speech and Language, a leading provider of speech & language technology. Cobalt can help your company achieve its potential as well. Contact Cobalt at info@cobaltspeech.com

Stas Tushinskiy, CEO and co-founder of Instreamatic, joins us in The Voice Box this week, as we explore how voice technology is helping people interact with their favorite brands through interactive ads. Instreamatic pioneered AI-driven voice advertising and is the leader in the voice ad tech space. Based on AI and deep learning, Instreamatic provides media companies and advertisers with a platform for interactive voice advertising. Instreamatic has established itself as an industry leader in the burgeoning field of Voice Advertising with Pandora, Salem Media, Gaana, Mindshare, Dentsu, Publicis and Hakuhodo being some of its clients.  Interesting Links: If you enjoyed listening to Stas, you can get more at his podcast, Voice AI in Marketing. They also have an associated LinkedIn community. --- The Voice Box is sponsored by Cobalt Speech and Language, a leading provider of speech & language technology. Cobalt can help your company achieve its potential as well. Contact Cobalt at info@cobaltspeech.com

David Bradford joins us in The Voice Box this week, as we explore how voice technology is helping people learn new languages. With FluentWorlds' language learning app, students can learn to speak English by immersing themselves in a virtual world where they interact with native English speakers. The app's Perfect Pronunciation tool gives users immediate feedback on how well they are pronouncing the words they are learning. Interesting Links: www.fluentworlds.com Fluentworlds Academy: https://academy.fluentworlds.com/ --- The Voice Box is sponsored by Cobalt Speech and Language, a leading provider of speech & language technology. Cobalt can help your company achieve its potential as well. Contact Cobalt at info@cobaltspeech.com

Have you ever had a bad experience with a call center agent? John Kane tells us what he and Cogito are doing to help coach agents and monitor calls for quality. John has a PhD in speech processing from Trinity College Dublin, which focused on developing novel techniques for analyzing voice quality. At Cogito, he runs the machine learning department. Please visit the Cogito engineering blog to learn more about the technology we discuss in today's episode. The Voice Box is sponsored by Cobalt Speech and Language, premier providers of speech technology to hundreds of companies large and small. Contact us.