Julian Quevedo 🌎

Hello!

I'm Julian, a 4th year undergrad at Stanford University, where I study computer science. I'm broadly interested in how we might create systems that match both the efficiency and capabilities of the human brain:)

Currently, I'm working on interactive video generation at a chip startup called Etched. Before that, I worked as a kernels engineer at Cohere and MosaicML to reduce the memory footprint of serving large GPT models. I was the first hire at Chord, an internet research agent (acq. by Newsweek). And in a past life, I worked with the Hogan Lab on an atom interferometer for detecting gravitational waves.

These days, I'm most excited about the future of world modeling, the intersection of neuroscience and AI, end-to-end robotics models, and full-stack ML systems optimization. I'm also trying to learn more about the history of computing and research organizations. In my free time I love to go on walks, and sometimes I draw with code. If you'd like to talk about any of these topics, or anything else at all, I'd love to hear from you!

github | twitter