AI SYSTEMS ENGINEER · SAN FRANCISCO · U.S. CITIZEN
I architect and operate production AI systems across agents, data, infrastructure, security, and product.
My work starts with ambiguous requirements and ends with a live system: measured, recoverable, and owned after launch.
I am most useful where model behavior, distributed state, security, and product constraints have to resolve into one production system.
AI Engineer · Production legal AI
Building and operating production legal-AI systems across document intelligence, agent infrastructure, evaluation, reliability, security, and observability.
Founding Engineer · Multimodal systems + 3D design
Built the AI catalog and discovery stack end to end—from heterogeneous supplier sites and render-ready materials to multimodal retrieval and the production design agent.
Public software and a paid consumer product: distribution, maintenance, economics, and user behavior included.
An open-source workspace orchestrator for running coding agents across isolated Git clones and persistent terminal sessions.
A live subscription iOS product whose product, native client, backend, AI behavior, payments, distribution, growth, and unit economics I own end to end.
A conceptual framework separating four places an AI system can fail: the human goal, the training objective, a learned internal objective, and the system’s effects on people.
Architecture. Implementation. Cutover. Recovery. Operation.