Notes

2026

Jul 03, 2026 · 6 min read

Lessons from building AI employee systems where reliability, tool boundaries, and observability matter more than prompt cleverness.

Apr 16, 2026 · 4 min read

Hallucinations, structured outputs, retries, validation, and human review are not edge cases; they are part of the architecture.

Nov 06, 2025 · 4 min read

AI applications need ordinary backend discipline: clear boundaries, asynchronous work, durable state, and integrations that can fail.

Aug 21, 2025 · 3 min read

A short note on event-driven systems, order books, latency budgets, streaming data, and why market systems make architecture concrete.

Oct 09, 2024 · 3 min read

Engineering lessons from building production systems when speed, ambiguity, and maintainability all matter.

Jun 17, 2024 · 1 min read

A one-minute note on why tool contracts matter more than giving an agent more tools.