I'm John. I work on AI systems — the practical engineering underneath model behavior: evaluation, tool use, agent loops, the places where probability meets production code.
This is where I write about it. Mostly notes from things that surprised me, occasionally longer pieces when something is worth working out in public.
Recent writing
- 06 Runway: capacity planning with calibrated uncertainty
- 05 Cairn: ranked attribution, not root cause
- 04 Anvil: the deploy gate run backward
- 03 Tessera: same engine, different scope, and what Anchor-from-round-1 looks like
- 02 DeploySignal: a statistical deploy gate, and the methodology that fell out of building one