Talk Python To Me #533: Web Frameworks in Prod by Their Creators
47 snips
Jan 5, 2026 
Join experts from the Python web framework universe: Carlton Gibson (Django core contributor), Sebastián Ramírez (creator of FastAPI), David Lord (lead maintainer of Flask), Janek Nouvertné, and Cody Fincher (Litestar maintainers). They dive into deploying applications in production, discussing everything from optimal database choices, like when to use SQLite versus Postgres, to deployment patterns with Kubernetes and Docker. Expect insights on avoiding blocking in async code, the performance boosts of upgrading Python, and deploying efficiently with FastAPI Cloud.
AI Snips
Chapters
Transcript
Episode notes
Pallets Site Uses SQLite For Speed
- David Lord runs the Pallets website using SQLite for static markdown content loaded at runtime.
- This approach gives very fast queries and a simple deployment for that site.
Let Kubernetes Do The Scaling Math
- Kubernetes is convenient for scaling by pod CPU/memory rather than manual process math.
- Keep pods small (1–2 CPUs) so the orchestration can autoscale effectively with traffic.
Start Small With Docker Containers
- Start small: deploy a single Docker container and scale only when needed.
- Use published Docker optimization guides to keep images tiny and reproducible for easy duplication.