Riddhish Bhalodia

I am a Research Scientist at Luma AI, working on post-training of video generation AI. Previously, I was a Research Scientist at Meta Superintelligence Labs, where I worked on LLM post-training and tool-calling, and contributed to the LLaMA family of models. Prior to that I was working in Meta Reality Labs, where I worked on codec avatars and expression tracking.

I completed my PhD in Computing in 2022 at Scientific Computing and Imaging Institute (SCI), University of Utah supervised by Ross Whitaker, Ladislav Kavan and Shireen Elhabian. I worked on healthcare AI and 3D medical vision for my dissertation. I earned my Bachelors in Electrical Engineering at Indian Institute of Technology, Mumbai in 2016.

My interests broadly lie in applied machine learnign and AI with an emphasis on 3D graphics and vision. I love building things that solve interesting real-world problems.

Outside work, you can usually find me reading a book while listening to a record. I also love boardgaming and hiking.