Talk Python To Me

Sydney Runkle, a LangChain developer who builds open source agent tooling, walks through Deep Agents and how they make models plan, iterate, and recover. She discusses planning and to-do systems, file system context, sub-agents and parallel work, middleware hooks, MCP interoperability, and using any model or local runtimes. Practical demos, CLIs, and safety sandboxes round out the conversation.

Martin Donath, open source developer behind Material for MkDocs and creator of Zensical. He explains why he rebuilt docs tooling into Zensical, a modern static site generator with a Rust core. They cover differential builds that run in milliseconds, a modular component-first design, migration tooling from MkDocs, and plans for search, CommonMark compatibility, and sustainable funding.

Samuel Colvin, creator of Pydantic and now leading Pydantic AI and Monty, builds a Rust-based sandboxed Python runtime for running LLM-generated code. He discusses Monty’s microsecond startup, strict sandboxing with host-mediated access, serializing interpreter state for resumability, and tradeoffs around limited Python features and no pip installs to improve safety and performance.

Amogh Desai, senior software engineer working on Airflow core and monorepo tooling. Jarek Potiuk, Apache Airflow maintainer and packaging expert. They tour a million-line Python monorepo. They discuss modern packaging PEPs, uv workspaces and uv sync for isolated per-package environments. They explain prek as a workspace-aware replacement for pre-commit and symlink-based shared libraries for modular architecture.

Carl Meyer, a Rust-based type checker developer with experience adding typing at Instagram. Rebecca Chen, a Python typing engineer who worked on Google’s Python team. Jelle Zijlstra, a typing council member focused on language governance at OpenAI. They explore how the typing council governs spec decisions, clashes between type checkers, new Rust-based checkers, nullability debates, and practical guidance on how much typing to add.

David Flood, a digital humanities developer on Harvard’s DARTH team, builds web platforms, archives, and tooling for research. He talks about keeping grant-funded sites alive by baking dynamic apps into static sites. He dives into client-side search with PageFind, running tooling in the browser with WebAssembly, and practical strategies for long-term archival and reuse.

Chris May, a Django developer and advocate who demos Datastar apps. Ben Croker, core maintainer focused on SDKs and docs. Delaney Gillilan, creator with a background in game and performance engineering. They discuss a tiny backend-driven framework that replaces HTMX+Alpine, server-sent-events for real-time streaming, multiplayer collaborative state, ultra-low-latency performance, and the design/philosophy behind Datastar.

Savannah Ostrowski, a Python Steering Council member who works on CPython internals and releases, and Sebastián Ramírez, creator of FastAPI and founder of FastAPI Cloud, talk about simplifying deployment. They discuss one-command deploys, autoscaling and scale-to-zero, and designing a Pythonic cloud with tooling that strengthens the open-source project.

Larry Ogrodnek, software developer and creator of PyView, brings Phoenix LiveView style real-time patterns to Python. He explains server-managed state with WebSocket diffs and how templates map to render trees. They explore event handling, JS hooks for rich front-end libs, T-Strings and Pythonic APIs, and deployment and scaling trade-offs for real-time Python web apps.

Vincent Warmerdam, a skilled data scientist and developer at Marimo, delves into the world of caching with DiskCache. He discusses the significance of caching in ML workflows, explaining how it saves time and costs. The conversation explores comparing traditional systems like Redis to SQLite’s innovative usage for persistent caches. Vincent shares insights on advanced features like sharding, memoization, and custom serialization, providing practical advice on designing effective cache keys. Tune in for practical tips and the future of caching in Python.