I build AI systems for teams that need to trust what those systems do.
I've spent years shipping products end to end, across mobile, web, and backend. I now lead technology as a CTO and build open-source tools for working with agents. Most of that work is in the constraints: what an agent is allowed to do, when a person approves it, and how you find out afterwards what happened.
What I focus on
- Agent systems with explicit tools, memory, approvals, and audit trails
- Workflow automation that improves execution without hiding operational risk
- Software architecture that teams can understand and evolve
- Developer tools built around clear interfaces and portable data
Public work
My current open-source work lives on GitHub. I publish field notes and case studies here, and share professional updates on LinkedIn.