About
Building AI systems.
I'm a lead AI/ML engineer focused on agents, automation, memory, and decision support. Clear system models, useful software. Build fast, break fast, learn fast.
Approach
What I spend time on
Most of my attention goes to the layer where model capability meets software structure and real production constraints.
Agents and control flow
How the inner loop is structured, what the harness does, and where the real system boundaries are.
Workflow automation
Using AI to remove bottlenecks, compress decision cycles, and automate the work worth automating.
Memory and decision support
Designing systems that retain relevant state, surface the right context, and help humans make better calls.
Working principles
Clear models, real constraints, and useful software.
I care about ideas that hold up under real workflows and business pressure — not just clean abstractions on paper. That means clear system models that make implementation easier, projects grounded in actual constraints, and writing that earns trust by being mechanically true. I avoid AI theatre dressed up as architecture, trend-chasing without implementation depth, and systems that feel clever but don't improve real work.