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)

Abstraction, but for robots

Abstraction, but for robots

Ryan welcomes Simone Kalmakis, VP of Engineering at Viam, to dive into how her team is bridging the gap between software and robotics, the importance of abstraction layers in making robotics more accessible, and the real-world applications of robotics from lobster traps to industrial sanding robots.Episode notes:Viam is a robotics platform that brings modern software development tools into hardware applications. Connect with Simone on Linkedin. This week’s shoutout goes to Lifejacket winner Sergey Kalinichenko for their answer to How does this K&R code for reading an int work?.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

2 Des 24min

Lightning-as-a-service for agriculture

Lightning-as-a-service for agriculture

Darryl Lyons, co-founder and Chief Rainmaker at Rainstick, joins the show to dive into advancements in AgTech and how Rainstick is using bioelectricity to enhance agricultural productivity. They discuss how Rainstick mimics natural thunderstorms to create electric fields and frequencies that promote plant growth, challenges and breakthroughs in their research, and their participation in the AWS Compute for Climate Fellowship.Episode notes:Rainstick uses electricity to mimic the natural effects of lightning to grow crops bigger, faster, and more sustainably. Want to learn more about the Compute for Climate program? Check our podcast with Lisbeth Kaufman, Head of Climate Tech at AWS.Ryan wrote about how software is being applied to agriculture a few years ago. Connect with Darryl on LinkedIn.Congrats to Lifeboat badge winner WestCoastProjects for their answer to Test accuracy is greater than train accuracy what to do?.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

28 Nov 21min

You’re probably underutilizing your GPUs

You’re probably underutilizing your GPUs

Ryan is joined by Jared Quincy Davis, CEO and co-founder of Mithril, to explore the importance of efficient resource allocation and GPU utilization in AI, the myth and misconceptions of the GPU shortage, and how the economics of GPU will change with new scheduling and utilization strategies. Episode notes:Mithril’s omnicloud platform aggregates and orchestrates multi-cloud GPUs, CPUs, and storage so you can access your infrastructure through a single platform.Connect with Jared on Twitter and LinkedIn.Shoutout to user Razzi Abuissa for winning a Populist badge on their answer to How to find last merge in git?.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

25 Nov 26min

Only you can stop AI database drops

Only you can stop AI database drops

Ryan is joined by David Hsu, CEO and founder of Retool, to explore how AI is transforming the role of a software developer into a software architect, the increasing accessibility of coding for non-engineers, and the importance of placing guardrails and higher-level programming primitives on AI coding assistants.Episode notes:Retool is an enterprise AI AppGen platform for internal software development, allowing you to create apps, agents, and workflows with any LLM, datasource, or API.David Hsu previously joined the podcast to complain about maintaining internal tools. Connect with David on Twitter.Today’s shoutout is for user Waseem Wisa, who won a Populist badge on their answer to Sass compiler throws 'undefined mixin' error when mixins are kept in seperate folder.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

21 Nov 32min

How to create agents that people actually want to use

How to create agents that people actually want to use

Ryan welcomes Assaf Elovic, head of AI at monday.com, to discuss creating AI tools that users will actually adopt, how they created their Monday Sidekick agent with the user experience in mind, and the opportunities that AI creates for better productivity and more efficiency.Episode notes:monday.com is a work management platform that allows you to plan and execute work across departments with a unified, AI-first product suite.Connect with Assaf on LinkedIn and Twitter. Congrats to Populist badge winner Vilx-, who received the badge for their answer to How do you provide a default type for generics?.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

18 Nov 27min

The fastest agent in the race has the best evals

The fastest agent in the race has the best evals

Ryan welcomes Benjamin Klieger, lead engineer at Groq, to explore the infrastructure behind AI agents, how you can turn a one-minute agent into a ten-second agent, and how they used fast inference and effective evals to build their efficient and reliable Compound agent. Episode notes: Groq delivers fast, low-cost inference using their custom-designed LPU, the first chip built for inference. Check out their agent, Compound, which can search the web and run code.Connect with Benjamin on LinkedIn and X. Congrats to user Bart Kiers for winning a Stellar Answer badge on their response to Regular expression to match a line that doesn't contain a word. See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

14 Nov 32min

One thing enterprise AI projects need to succeed? Community.

One thing enterprise AI projects need to succeed? Community.

In this episode of Leaders of Code, Stack Overflow CEO Prashanth Chandrasekar chats with Ramprasad Rai, VP of Platform Engineering at JPMorgan Chase & Co., about the unique challenges of implementing AI in an enterprise environment. They discuss how organizations can balance AI-driven productivity with strict compliance and security requirements by leveraging a community-driven knowledge system that grounds probabilistic AI tools in internal, trusted expertise.The discussion also:Explores why AI models often hallucinate in enterprise environments due to a lack of internal context. Highlights how Stack Overflow’s structured Q&A data provides ideal fine-tuning material for the next generation of AI models.NotesConnect with Ramprasad Rai on LinkedIn.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

13 Nov 23min

AI code means more critical thinking, not less

AI code means more critical thinking, not less

Ryan is joined by Secure Code Warrior’s co-founder and CTO Matias Madou to discuss the implications of LLMs’ variability on code security, the future of developer training as AI coding assistants become more popular, and the importance of critical thinking—especially for junior developers—in the age of AI.Episode notes: Secure Code Warrior upskills development teams to help companies stay protected against potential cybersecurity threats.Connect with Matias on Linkedin. Shoutout to Lifejacket badge winner Sergey Kalinichenko, who won the badge for their answer to K&R Code for getting an int.TRANSCRIPTSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

11 Nov 32min

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