Welcome

This is the place where the practice’s notes will live — the patterns, opinions, and lessons that surface from building production AI systems with real teams.

Expect short pieces, not long ones. The bias here is toward writing things down only when they are useful to someone else trying to make a decision: how to scope an eval harness, when RAG is the wrong tool, why most agent failures are actually just bad observability, and so on.

The home page lists the engagement shapes and the services. This section will fill in over time.