FIELD NOTE / VLM Evaluation
Claude, chessboards, and spatial reasoning
A small chessboard-to-FEN experiment showed where Claude vision struggles: not global understanding, but exact square-level localization.
Read the noteEngineering from the team building VideoDB.
Field notes, practical build stories, and newsletter issues from the work behind VideoDB.
featured
field notes
Operations
Reuse HTTP connections in Python services
A short field note on replacing repeated bare requests calls with a shared session so Python services can reuse HTTP connections.
Frontend Performance
One CORS header that removed hundreds of milliseconds from repeated API calls
A practical note on preflight caching, custom auth headers, and why OPTIONS requests quietly dominate app latency on slower connections.
Operations
A Postgres backup is not real until you restore it
A field note from a simple offline Postgres migration: why pg_dump was enough, where read-only mode fits, and why the restore drill matters more than the dump command.
how i built
How I Built
Deep Search: How We Built an Engine for Finding Exact Moments in Video
How we built Deep Search as a retrieval loop for finding exact moments in video using planning, indexing, validation, recovery, and follow-up state.
How I Built
Pair Programmer: live screen and audio context for coding agents
A short engineering note on turning a developer session into searchable context with VideoDB Capture, RTStreams, local events, and agent-side retrieval.
How I Built
How I built CallMD: AI call intelligence on top of VideoDB
A practical build story for a local-first call intelligence app that records calls, transcribes speakers, generates live nudges, and exports structured Markdown.
newsletter
Newsletter / Video agents · model watch · VideoDB updates
The VideoDB Dispatch #001
Why agent runs need playable evidence, thinking tokens in vision models and agentic video streams.
subscribe