Building AI-native products with agents, harnesses, and tight feedback loops.
I build products, not just prototypes.
My work sits at the intersection of AI product development, agentic tooling, and automation systems. I care about clear goals, fast MVP convergence, strong verification loops, and turning ideas into products people can actually use.
- AI products for real workflows
- agentic tooling for execution and repair
- automation systems that reduce repeat work
- Humans define goals. AI executes.
- Converge the MVP before building.
- Use harnesses to control quality.
- Verify before handoff.
- Ship, learn, repair, repeat.
- building AI-native products with clear user value
- exploring harness-first agentic development
- turning product ideas into testable MVPs fast
|
|
|
My default operating model is simple:
humans steer, agents execute, harnesses control quality