Hey — thanks for finding your way here.

These days I'm building in physical AI. I'm Founding GTM at PrismaX — we provide the real-world teleoperation demonstration data that robot foundation models (VLA) and world models learn from. Robots are hitting the same wall LLMs once did: the models are ready, but the data that makes them reliable in the physical world is the bottleneck. That's the layer we're building.

Before that, I spent 5 years as an early-stage investor at Symbolic Capital — 30+ investments across AI and blockchain. That taught me a lot about what separates real products from vaporware. Eventually I decided I'd rather build than bet.

Background: Math (Peking University), quant modeling (Texas A&M), MBA (Wharton). The common thread is curiosity — it's gotten me into some interesting places.