AI SYSTEMS FOR THE REAL WORLD
Agentic AI, minus the theater.
I lead agentic architecture, enterprise AI platforms, evaluation, observability, secure tool use, and governance that survives contact with reality.
The useful detour
Built from systems thinking.
- Analytics Find the signal.
- Cloud Make it scale.
- Infrastructure Keep it standing.
- Applied AI Now add judgment.
A quick tour
The whole site, without the whole speech.
What the demo forgot.
Essays and quick lessons from building AI systems in the real world.
Read the notes → ProjectsProof, with tradeoffs.
Working builds, architecture decisions, GitHub links, and honest constraints.
View projects → AboutHuman, still included.
The short version of my work, credentials, controlled chaos, and how to reach me.
About and contact →Currently useful
Fresh thinking. Working code.
Agents With Brakes: Production Agents Need Control Systems
A detailed write-up on Agents With Brakes, a reference architecture for enterprise agent platforms built around policy, identity, tool access, evals, observability, approvals, cost control, memory, and failure management.
Featured projectAgents With Brakes
A public reference architecture for production-grade enterprise agent platforms with policy, identity, evals, observability, approvals, rollback paths, and failure management.