I work at the level of the whole system.
I am an AI systems engineer based in San Francisco. I have led or originated systems spanning live legal intake, multimodal catalog generation, distributed workflows, customer security, observability, open source, and paid consumer software.
How I work
- 01
Own the full arc
Architecture is incomplete until the migration, rollout, failure handling, and operating model are real.
- 02
Make failure explicit
Retries, identity, provenance, access, and recovery belong in types, schemas, and durable state—not tribal knowledge.
- 03
Evaluate the real task
I distrust clean benchmarks until the inputs, replay path, and ground truth match production behavior.
- 04
Stay close to the product
Latency, cost, security, and model quality only matter through the workflow a customer actually experiences.
I was born in Ukraine, moved to the United States as a child, entered college at fifteen, and became a U.S. citizen. I studied philosophy, logic, computation, and AI ethics before making engineering my full-time practice.
The academic background is useful because it trained precision around goals, evidence, and causality. The engineering record stands on its own: production systems, public software, operating outcomes, and responsibility after launch.
Research and educationSan Francisco, CA · U.S. citizen