I am a research lead at Luma AI, currently leading pre-training and post-training for large-scale, generalizable world action models. Previously, I led pre-training and post-training for Luma’s UNI-1 multimodal image model, as well as follow-up research efforts for UNI storyboard (Luma Scenes), manga, and super-resolution.
I was also the CTO and co-founder of Apparate Labs (acquired by Luma AI), where we trained a real-time video foundation model for generating highly realistic and expressive humans.
I received my PhD in computer science from Stanford University, co-advised by Leonidas Guibas and Gordon Wetzstein. During my PhD, I interned at Google Research, NVIDIA Research, and Adobe Research.
Previously, I worked at Google as a software engineer on Portrait Mode. I was also a Computer Science undergraduate and graduate student at Carnegie Mellon University, advised by Keenan Crane.
Outside of work, I enjoy cooking, badminton, swimming, boardgames, and music (long-time swiftie). Also a fan of traveling and sports – go Ravens!
Interests
- Computer Vision
- Computer Graphics
- Generative AI
- Geometry Processing
Education
-
PhD in Computer Science, 2020-2024
Stanford University
-
MSc in Computer Science, 2018-2019
Carnegie Mellon University
-
BSc in Computer Science, 2015-2018
Carnegie Mellon University