Blog
Browse by category or scroll for the latest posts.
A short hello and what to expect from this newsletter.
Most engineers only realise it after a retry storm takes down a production system.
A walkthrough of the multi-stage pipeline that turns raw bytes into a queryable knowledge base — from Spanner-backed queues to Gemini's hybrid grounding approach.
A clear mental model for picking the right AI architecture.
There is no perfect backend architecture. Only systems that match their context — and systems that don't.