The Stack Overflow Podcast

Jon Hyman, co-founder and CTO of Braze, a leader in engineering and AI-first product development. He recounts rapid AI adoption, building AI-powered prototyping and tooling, and shifting team structures for scale. He discusses winning skeptics with model quality, managing inference costs at scale, and preparing engineering orgs and agents for autonomous feature work.

Philip Rathle, CTO at Neo4j, a leader in graph databases and knowledge layers for AI. He explains why model-only approaches struggle in enterprises. He describes Graph RAG: combining knowledge graphs with vectors for targeted, connected context. He discusses avoiding context rot, deterministic multi-hop reasoning, graph embeddings, and real-world production uses like Uber and Walmart.

Mark Papermaster, AMD CTO with deep expertise in CPU/GPU architectures, discusses AMD’s silicon strategy rooted in heterogeneous computing. He covers chiplet design and manufacturing agility. He talks about tailoring chips for varied AI workloads, software stacks like ROCm, Helios rack/system optimization, and how agentic AI both consumes and accelerates chip innovation.

Brian O'Grady, Head of Field Research and Solutions Architecture at Qdrant, specializes in vector search, embeddings, and scalable search systems. He contrasts traditional Lucene text search with modern vector databases. He explains when exact-match search is preferable, the trade-offs of bolt-on vector indexes, Qdrant's portability and scaling features, and future work on video embeddings and local agent contexts.

Jason Williams, senior software engineer and creator of the Rust JS engine Boa, and TC39 delegate focused on Temporal. He discusses why date and time are so hard in JavaScript. He explains Date mutability problems and why libraries like Moment.js fell short. He describes Temporal’s types, precision choices, and the long standardization process.

Harsha Chintalapani, Co-founder and CTO at Collate and metadata expert who built large-scale data systems, talks about why AI struggles with real-time, structured production data. He discusses schema drift, inconsistent definitions like “customer,” and weak governance. The conversation covers metadata, observability, knowledge graphs, and practical first steps teams can take to get data AI-ready.

Josiah McGarvey, filmmaker and co-owner at Cult.Repo who makes documentaries about open-source communities. Emma Tracy, co-founder and producer at Cult.Repo who crafts human-centered tech stories. They discuss making films about open-source people, sourcing interviews within tight communities, balancing technical depth with narrative, sustainability and burnout in projects, and upcoming documentaries including an AI anthology.

Steven Kulesza, staff software engineer at Intuit focused on multi-agent orchestration and safe AI in finance. Chase Ruzin, engineering manager building AI-enabled product experiences and orchestration layers. They discuss coordinating many specialized agents, evaluation strategies for multi-agent systems, when humans should review outputs, ensuring numeric correctness in finance, central planner vs agent swarms, and operational cost and observability challenges.

Mike Swift, co-founder and CEO of Major League Hacking, leads a global community that runs hackathons and educational programs. He discusses why developer communities and entry points into programming still matter. He talks about MLH’s acquisition of DEV and how they’re building shared spaces for learning and publishing. He also explores how AI is changing who can create and how communities will evolve.

Stephen Watt, distinguished engineer and VP at Red Hat’s Office of the CTO, brings expertise in distributed systems, Kubernetes, and the PyTorch ecosystem. He unpacks digital sovereignty, infrastructure limits like power and cooling, and why regional constraints shape sovereign AI. He also explores extending Kubernetes for AI stacks, modular small models, semantic routing, and the push for open silicon and on-prem strategies.