I started by cutting my teeth building distributed systems (like Wormhole and the Distributed Cloud Operating System). At the same time, Ethereum was beginning to surface as something uniquely ambitious: a programmable global substrate with technical, economic, game-theoretic, legal, and governance implications. I went all in.

I contributed to early Ethereum R&D, including State Channels - an experimental scaling technology. We were building the infrastructure that would've enabled micropayments and agentic systems - too early for its time, but frontier work.

From there, I took a break from building blockchain systems and shifted into applied AI. I set out to build recommender systems, and luxury e-commerce (at Farfetch) provided that ground, where recommendations are a part of the product UX. But while building recommender systems, I kept having conversations around the nature of crypto systems and what still needed to be built in the long term.

So I joined Polychain Capital, an early-stage crypto-focused venture capital firm. As a research and investment partner, I primarily focused on infrastructure, AI, and consumer.

Research

Early on I co-authored a paper with Ertem Nusret Tas (Stanford) et al, specifying the formal security properties of another Ethereum scaling technology called "rollups or L2s." This technology ended up being the foundation of Ethereum's scaling roadmap.

Investments

Across various investments that I led or diligenced, CreateSafe is a notable one. CreateSafe was at the frontier of music, verifiable ownership (using blockchain rails), and AI. Their platform enabled Grimes to upload her voice as an AI model, her fans to make songs with her voice, and for revenue to be split between fans and Grimes. Time recognized Grimes as a leader shaping the future of AI, while Marc Andreessen and Ben Horowitz called this out as "one of the most innovative things someone's done in tech." CreateSafe may also have been on the earlier side :)

From Polychain I turned my attention to building again, this time at EigenCloud, at the intersection of blockchains and AI. I started out as a Principal Research Engineer then transitioned into AI General Manager where I led the launch of EigenAI - a deterministic LLM API (more details here: here).

This determinstic LLM API enables inference to be verified (through re-execution), which is needed in high-stakes applications like agentic finance or LLM auditability / compliance.