I'm a researcher at Genmo, working on world foundation models for robots. My work sits between generative modelling and the systems that make it scale: I build large-scale image and video models, and usually own the training stack alongside them.
Before Genmo I worked on video generation at Leonardo AI (acquired by Canva) and Haiper, and on parameter-efficient training of LLMs at Huawei's Noah's Ark Lab. I did my MSc at Mila, advised by Eugene Belilovsky, on continual and decentralized learning.
Always happy to chat about research, startups, and new ideas — feel free to reach out.
The latest
- April 2025 Joined Genmo.ai as a Member of Technical Staff.
- March 2025 Shipped Leonardo AI Motion 2.0.
- Oct 2024 Shipped Haiper 2.0.
Products
-
Haiper 2.0 Haiper
Second-generation text-to-video model, with sharper output and more controllable motion.