About

I trained in pure mathematics at Moscow State University, studied law at the London School of Economics, and spent five years in London finance before moving to San Francisco to build companies. The thread connecting these isn't ambition or restlessness—it's a persistent interest in how systems behave: legal systems, mathematical structures, financial instruments, AI architectures. Each domain has its own formalism for describing what's possible, what's constrained, and where the structure breaks down.

I've founded and exited an AI-driven fintech, led product at Twitter's Cortex team during a period when algorithmic accountability was moving from academic concern to operational reality, and co-led a civilian AI deployment in an active conflict zone. Most recently, I've been working on constitutional AI standards with the Hong Kong Monetary Authority and developing frameworks for CBRN information hazards in multi-agent systems.

My current orientation is toward AI safety as a continuous practice rather than a destination—operational risk, systemic risk, and the consciousness constraints that arise when intelligence lacks embodied experience. I'm interested in what constitutional AI shares with constitutional law, why general language shapes model behavior more effectively than prescriptive rules, and what continental philosophy can contribute to alignment research that analytic philosophy hasn't already provided.