3 field notes · from the build
Notes from the build.
Field notes on engineering AI systems that actually ship — agents, evals, retrieval, observability, and deploy pipelines. What's working in production, and what we learned the hard way.
What we write about
01
Evals & Testing
Why we write the evals before the prompts — measuring quality before anything ships.
02
Observability
What it actually means to see inside an AI agent once it's running in production.
03
RAG & Retrieval
Building retrieval that stays accurate and trustworthy as the corpus grows.
04
Agents & Automation
Designing agents that do real, reliable work — not demos that fall over.
