I build tools that sit at the boundary between "this should just work" and "this needs a human to look at it" — rules engines, safety systems, and the occasional field guide to the mistakes engineers make under pressure.
- 🔭 Currently maintaining osprey — a high-performance safety rules engine for real-time event processing at scale.
- 📖 Also wrote cognitive-biases-in-software-engineering — concrete SWE scenarios for the biases that quietly wreck estimation, review, debugging, and incident response.
- 🌱 Interested in systems that fail loudly instead of silently, and in the small process changes that prevent the same incident from happening twice.
- ⚡ Fun fact: this avatar is generated, not drawn — a small Pillow script produces the gradient, glow, and monogram from scratch, no assets.
- Prefer boring, provable infrastructure over clever, fragile infrastructure.
- A rule that nobody can explain is a rule that will misfire in production.
- Write the postmortem before you need one: name the failure modes early.
Thanks for stopping by.