Your LLM issues are really data issues

Your LLM issues are really data issues

Ryan welcomes Harsha Chintalapani, co-founder and CTO at Collate and co-creator of Open Metadata, to the show to discuss why AI and LLMs struggle with real-time, structured production data. They explore how schema changes, inconsistent definitions (like “customer”), and weak governance can break both your analytics and MLs, and what companies can do to get their data AI-ready, from metadata management to observability.  Episode Notes:  Collate [https://www.getcollate.io/] is a semantic intelligence platform built on a semantic metadata graph for discovery, governance, and AI observability across your data ecosystem. Connect with Harsha on LinkedIn [https://www.linkedin.com/in/sriharsha/].  Congrats to user buttonsrtoys [https://stackoverflow.com/users/2079612/buttonsrtoys], who won a Famous Question badge for their question Possible to edit PDF without embedded font installed? [https://stackoverflow.com/questions/27807875/possible-to-edit-pdf-without-embedded-font-installed]. See Privacy Policy at https://art19.com/privacy [https://art19.com/privacy] and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info [https://art19.com/privacy#do-not-sell-my-info].

Lights, camera, open source!

Ryan is joined on the show by Cult.Repo producers Emma Tracey and Josiah Mcgarvie to discuss making documentaries about open-source software and the people behind the major technologies that uphold the internet. They explore why open-source projects and the people who maintain them are such interesting stories for audiences, how being outsiders has helped them tell these community stories, and what they see as the common stressors that plague all open-source projects, such as sustainability, compensation, and burnout.  Episode notes:  Cult.Repo [https://www.cultrepo.com/] produces documentaries and shorts about the human stories behind open-source technology. Check out their filmography on their YouTube channel [https://www.youtube.com/channel/UCsUalyRg43M8D60mtHe6YcA]. Have an idea for an open-source community they should cover? Email the Cult.Repo team at hello@cultrepo.com.  Shoutout to user kiranvj [https://stackoverflow.com/users/1188322/kiranvj] for winning a Populist badge for their answer to What is a good way to automatically bind JS class methods? [https://stackoverflow.com/questions/56503531/what-is-a-good-way-to-automatically-bind-js-class-methods]. TRANSCRIPT [https://stackoverflow.blog/2026/04/24/lights-camera-open-source/] See Privacy Policy at https://art19.com/privacy [https://art19.com/privacy] and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info [https://art19.com/privacy#do-not-sell-my-info].

How to get multiple agents to play nice at scale

SPONSORED BY INTUIT Chase Roossin, group engineering manager, and Steven Kulesza, staff software engineer, from Intuit join the podcast to chat about what might be the hardest problem in engineering right now: getting multiple AI agents to work together in a complex system. They discuss how automated evals can make agent behaviors more predictable, agent swarms vs. one highly skilled agent, and how customer behavior shaped their technical architecture.  Episode notes Want to work on complex engineering problems like these? Explore careers [https://www.intuit.com/careers/?cid=dis_so_clicks_us_intuit-intelligence_aw_fy26-pod_alltechaudience_link_none_intuit-talent__] at Intuit. We’ve worked with Intuit on a few other great blogs and podcasts, including Best practices for building LLMs [https://stackoverflow.blog/2024/02/07/best-practices-for-building-llms/] and How Intuit democratizes AI development across teams through reusability [https://stackoverflow.blog/2023/03/01/how-intuit-democratizes-ai-development-across-teams-through-reusability/]. Connect with Chase on LinkedIn [https://www.linkedin.com/in/chaseroossin/].  Connect with Steven on LinkedIn [https://www.linkedin.com/in/steven-kulesza-27240540/].  Congrats to Lifejacket badge winner Sean [https://stackoverflow.com/users/5351721/sean] for saving Creating the simplest HTML toggle button? [https://stackoverflow.com/questions/76837048/creating-the-simplest-html-toggle-button/76837247#76837247] with a great answer. TRANSCRIPT [https://stackoverflow.blog/2026/04/22/how-to-get-multiple-agents-to-play-nice-at-scale/] See Privacy Policy at https://art19.com/privacy [https://art19.com/privacy] and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info [https://art19.com/privacy#do-not-sell-my-info].

We still need developer communities

Ryan welcomes Mike Swift, co-founder and CEO of Major League Hacking, to the show to chat about the never-ending need for software developer communities and entry points into programming; MHL’s recent acquisition of DEV and how they’re creating a place for shared knowledge, building, and publishing; and why now is the best time to be both an artisan and a builder in a world with AI software development tools. Episode notes:  Major League Hacking [https://mlh.io/] is a 500k+ global member community that hosts hackathons and open-source fellowships for the next generation of developers. They recently acquired DEV [https://dev.to/], an online community for 3M+ developers to learn and share together.  Connect with Mike on LinkedIn [https://www.linkedin.com/in/theycallmeswift/] or email him at swift@mhl.io [swift@mhl.io].  Congrats to Stellar Answer badge winner Antony Hatchkins [https://stackoverflow.com/users/237105] for getting over a hundred saves on their answer to Git replacing LF with CRLF [https://stackoverflow.com/questions/1967370/git-replacing-lf-with-crlf].  TRANSCRIPT [https://stackoverflow.blog/2026/04/21/we-still-need-developer-communities/] See Privacy Policy at https://art19.com/privacy [https://art19.com/privacy] and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info [https://art19.com/privacy#do-not-sell-my-info].

No country left behind with sovereign AI

Ryan welcomes Stephen Watt, distinguished engineer and VP of Red Hat’s Office of the CTO, to chat about digital sovereignty and sovereign AI. They explore major infrastructure constraints for things like power, cooling, and scarce hardware that cause the regional disparities we see in sovereign AI, plus why we need to extend Kubernetes and integrate PyTorch Stack not just for a sovereign cloud but for sovereign AI. Episode notes:  Red Hat [http://redhat.com]’s Office of the CTO is a division of 150 software engineers and researchers working on their Research [https://research.redhat.com/] and Emerging Technologies [https://next.redhat.com/] arms, helping to shape the vision and strategy of Red Hat’s technology.  Connect with Stephen on LinkedIn [https://www.linkedin.com/in/wattsteve/].  Congrats to user Ittiel [https://stackoverflow.com/users/1355777/ittiel] for winning a Populist badge on their answer to Print timestamps in Docker Compose logs [https://stackoverflow.com/questions/30003464/print-timestamps-in-docker-compose-logs]. TRANSCRIPT [https://stackoverflow.blog/2026/04/17/no-country-left-behind-with-sovereign-ai/] See Privacy Policy at https://art19.com/privacy [https://art19.com/privacy] and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info [https://art19.com/privacy#do-not-sell-my-info].