I've spent the last decade at the intersection of cutting-edge machine learning research and real-world product impact. I've built AI systems used by billions of people, backed early-stage founders building the next generation of AI companies, and helped senior leaders figure out what this technology actually means for how they work and lead.
I build products, invest in companies, coach executives, and host communities. Each of those roles makes the others sharper. The common thread is a practitioner's relationship with AI: I've shipped it, studied it, bet on it, and taught it.
I studied Computer Science at Harvard, where I developed a deep interest in how intelligent systems perceive and make sense of the world. That curiosity drove everything that came after.
I co-founded a startup out of Y Combinator's W20 batch. Building from zero taught me things no amount of time at a big company could: how to move fast under uncertainty, how to make product decisions with almost no data, and what it actually feels like to be responsible for the whole thing.
I joined Google Brain as a Research PM, embedded in the research organization behind the AI systems powering Alphabet's most important products. I worked at the intersection of research and product, translating cutting-edge ML work into real systems that fed directly into products like Google Search, Pixel, and Waymo. One project from this period was recognized with a Cannes Lion award for authentic representation in imaging technology. I also published research at NeurIPS during this time.
Most recently, I was a PM Lead at Google DeepMind, working on Gemini, self-coding agents, self-evolving systems, and reinforcement learning research. This is some of the most technically complex and consequential work happening in AI today.
The most useful perspective on AI transformation comes from people who've actually shipped AI systems, not people who've only read about them.
The connections made in a small room with the right people change careers, companies, and sometimes entire fields.
No matter how impressive the model, the product only works if it fits into someone's life in a way that actually helps them.
Not as a buzzword. As a first-order product and research concern. Who builds the system shapes what the system sees and does.
Want to work together? I'm open to coaching, advising, and conversations with people building at the frontier.
Get in touch