Working notes from the team building production AI platforms — Atheneo, HomeServiceOne, CleanSlate — and advising the companies scaling their own. No thought-leadership fluff. Real answers to the questions operators actually ask.
Model routing, prompt caching, request batching, and per-account cost ceilings — the four disciplines that keep AI margins intact as usage scales. A field guide from the platforms we run ourselves.
Read the guide →Every piece is grounded in a live production system — ours or a client's.
The four cost-control disciplines every production AI system needs — model routing, caching, batching, and per-account ceilings.
Coming soon. A framework for scoring AI opportunities by impact × feasibility, drawn from real client engagements.
Coming soon. Category-by-category — what's essential, what's optional, and what to rip out.
Book a 30-minute strategy call. Every framework you read here is something we run on real engagements — happy to run it on yours.
Book a Strategy Call