From search trees to neural nets, a deep dive into natural language processing
The Stack Overflow Podcast23 Jun 2021

From search trees to neural nets, a deep dive into natural language processing

We chatted with three guests:

Miguel Jetté: Head of AI R&D

Josh Dong: AI Engineering Manager

Jenny Drexler: Senior Speech Scientist

When Jette was studying mathematics in the early 2000s, his focus was on computational biology, and more specifically, phylogenetic trees, and DNA sequences. He wanted to understand the evolution of certain traits and the forces that explain why our bones are a certain length or our brains a certain size. As it turned out, the algorithms and techniques he learned in this field mapped very well to the emerging discipline of automatic speech recognition, or ASR.

During this period, Montreal was emerging as a hotbed for artificial intelligence, and Jette found himself working for Nuance, the company behind the original implementation of Siri. That experience led him to several positions in the world of speech recognition, and he eventually landed at Rev, where he founded the company’s AI department.

Jette describes Rev as an “Uber for Transcription.” Anyone can sign up for the platform and earn money by listening to audio submitted by clients and transcribing the speech into text. This means the company has a tremendous dataset of raw audio that has been annotated by human beings and, in many cases, assessed a second time by the client. For someone looking to build an AI system that mastered the domain of speech to text, this was a goldmine.

Jette built the earliest version of Rev’s AI, but it was up to our second guest, Josh Dong, to productize and scale that system. He helped the department transition from older technologies like Perl to more popular languages like Python. He also focused on practical concerns like modularity and reusable components. To combine machine learning and DevOps, Dong added Docker containers and a testing pipeline. If you’re interested in the nuts and bolts of keeping a system like Rev’s running at tremendous scale, you’ll want to check out this part of the show.

We also explore some of the fascinating future and promise this technology holds in our time with Jenny Drexler. She explains how Rev is moving from a hybrid model—one that combines Jette’s older statistical techniques with Dong’s newer machine learning approach—to a new system that will be ML from end-to-end. This will open up the door for powerful applications, like a single system that can convert speech text across multiple languages in a single piece of audio.

“One of the things that's really cool about these end to end models is that basically, whatever data you have, it can learn to handle it. So a very similar architecture can do sequence to sequence learning with different kinds of sequences. The model architecture that you might use for speech recognition can actually look very similar to what you might use for translation. And you can use that same architecture, to say, feed in audio in lots of different languages and be able to do transcription for any of them within one model. It's much harder with the hybrid models to sort of put all the right pieces together to make that happen,” explains Drexler.

If you’re interested in learning more about the past, present, and future of artificial intelligence that can understand our spoken language and learn how to respond, check out the full episode. If you want to learn more about Rev or check out some of the positions they have open, you can find their careers page here.

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Episoder(896)

AI agents for your digital chores

AI agents for your digital chores

Ryan welcomes Dhruv Batra, co-founder and chief scientist at Yutori, to explore the future of AI agents, how AI usage is changing the way people interact with advertisements and the web as a whole, and the challenges that proactive AI agents may face when being integrated into workflows and personal internet use. Episode notes:Yutori is building AI agents that can reliably handle everyday digital tasks on your behalf on the web.Connect with Dhruv via his website. Congrats to the winner of today’s Populist badge, user Don Kirkby, who earned it with their answer to Find all references to an object in python.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

14 Okt 34min

Vite is like the United Nations of JavaScript

Vite is like the United Nations of JavaScript

Ryan welcomes back Evan You, the creator of Vite and Vue.js, to discuss the evolution of build tools in web development, the unique features of Vite from its plugins to its hot module capabilities, and the future of Vite, including its integration with Rust. Plus, they touch on Vite’s new documentary and the power of open-source communities.Episode notes:Vite is a frontend build tool powering the next generation of web applications. Check out all of the work Evan is doing at his company VoidZero. For more on the origins of Vite, watch the newly-released Cult.repo’s documentary. If you’d rather hear Evan talk about Vue.js, listen to the podcast we published with him earlier this summer. Today’s shoutout goes to user dbush for winning a Populist badge on their answer to How does printing a union itself and not its member work in C?.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

10 Okt 27min

Context is king for secure, AI-generated code

Context is king for secure, AI-generated code

Ryan sits down with Dimitri Stiliadis, CTO and co-founder of Endor Labs, to talk about how AppSec is evolving to address AI’s use cases. They discuss the implications of AI-generated code on security practices, the importance of human oversight in managing vulnerabilities, and how organizations should be balancing security and efficiency with AI. Episode notes:Endor Labs is AppSec for the software development revolution, helping you pinpoint critical risks whether your code is written by a human or AI. Connect with Dimitri on LinkedIn.Today’s shoutout is for user skovorodkin, whose answer for Elegant Python code for Integer Partitioning  was so good, it outscored the accepted answer. See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

7 Okt 28min

One is not the loneliest number for API calls

One is not the loneliest number for API calls

Gil Feig, co-founder and CTO of Merge, joins the show to explore Merge’s approach for reducing third-party APIs to a single call, the complexities of and need for data normalization, and the role that AI and MCP plays in the future of API functionality. Episode notes: Merge connects you to any third-party system for fast, secure integrations for your products and agents.Connect with Gil on LinkedIn and X. Shoutout to user Abhijit for winning a Lifeboat badge on their answer for Complex numbers in python.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

3 Okt 26min

Building AI-ready teams: Why documentation and culture matter more than tools

Building AI-ready teams: Why documentation and culture matter more than tools

In the second part of this two-part Leaders of Code episode, Peter O'Connor, Director of Platform Engineering, and Ryan J. Salva, Senior Director of Product at Google Developer Experiences, dive beyond AI hype to explore the shifts reshaping how engineering teams operate and scale. From the critical role of documentation quality in AI workflows to the cultural transformations needed for successful adoption, Peter and Ryan discuss the deeper implications of integrating AI into modern software development practices.The discussion also:Explores how poor documentation creates problems as AI systems learn and repeat mistakes, making high-quality documentation essential for successful AI integrations.Covers how consistent tools and processes become more important when using AI, and why leaders should prioritize helping teams learn and experiment with AI tools instead of just measuring productivity.Offers practical advice for leaders on how to create environments where developers can learn and build confidence with AI tools.Notes:Listen to Part 1 of the conversation here.Connect with Peter O’Connor and Ryan J Salva.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

2 Okt 20min

As your AI gets smarter, so must your API

As your AI gets smarter, so must your API

Ryan sits down with Marco Palladino, CTO of Kong, to talk about the rise of AI agents and their impact on API consumption, the MCP protocol as a new standard for agents, the importance of observability and security in AI systems, and the importance for businesses and entrepreneurs to leverage opportunities in the agentic AI space now. Episode notes: Kong is an all-in-one API platform for AI and agentic workflows.Marco previously joined the podcast in 2024. Connect with Marco on Twitter.Congratulations to user Mark for receiving a Lifeboat badge for their answer to Visual Studio Code: Expand the horizontal bar for scrolling tabs.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

30 Sep 28min

Getting Backstage in front of a shifting dev experience

Getting Backstage in front of a shifting dev experience

Ryan welcomes Pia Nilsson, GM for Backstage and head of developer experience at Spotify, to discuss the evolution and adoption of Backstage, the impact of AI on dev experience, and how Spotify approaches platform engineering and standardization to help teams solve for specific needs. Episode notes: Backstage is an open-source IDP by Spotify that reduces everyday friction, cognitive overhead, and operational toil for developers.We previously talked to the Backstage team in 2022.Poor Ryan. If only he had Backstage at his last job. You know what makes Backstage even better? Our Stack Overflow for Teams integration. Connect with Pia on LinkedIn.Markus Pscheidt gets today’s shoutout for winning the Populist badge with their answer to Dynamic tag values for the Counter metric in Micrometer.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

26 Sep 26min

Democratizing your data access with AI agents

Democratizing your data access with AI agents

Jeff Hollan, director of product at Snowflake, joins Ryan to discuss the role that data plays in making AI and AI agents better. Along the way, they discuss how a database leads to an AI platform, Snowflake’s new data marketplace, and the role data will play in AI agents. Episode notes:Snowflake provides a fully-managed data platform that developers can build AI apps on. We’re happy to have Stack Exchange data available on the Snowflake Marketplace. Connect with Jeff on LinkedIn and Twitter. Congrats to Timeless for throwing a Lifejacket to Using pandas to read HTML.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

23 Sep 29min

Populært innen Business og økonomi

stopp-verden
lydartikler-fra-aftenposten
dine-penger-pengeradet
e24-podden
rss-penger-polser-og-politikk
rss-borsmorgen-okonominyhetene
kommentarer-fra-aftenposten
pengepodden-2
finansredaksjonen
utbytte
pengesnakk
tid-er-penger-en-podcast-med-peter-warren
morgenkaffen-med-finansavisen
rss-markedspuls-2
stormkast-med-valebrokk-stordalen
livet-pa-veien-med-jan-erik-larssen
rss-investering-gjort-enkelt
rss-fri-kontantstrom
okonomiamatorene
paretopodden