I work at the intersection of physics, AI, and engineering. At NVIDIA, I'm a Senior Physics-Machine Learning Engineer on the PhysicsNeMo team, working on scalable physics-AI frameworks for scientific computing and engineering.
Before NVIDIA, I spent five years at Autodesk as a Principal AI Research Scientist, most recently in the Autodesk AI Lab, where I worked on large-scale training of vision-language models along with fine-tuning and distillation techniques. More broadly, I led research on generative and agentic AI for engineering design, spanning transformer-based models for mechanical assembly and kinematic mechanism synthesis, multi-agent LLM frameworks for design applications, and simulation-feedback fine-tuning of generative design models. During that time I also created XLB, a differentiable lattice-Boltzmann library in JAX and NVIDIA Warp.
I hold a Ph.D. in Computational Physics from the University of Toronto and was a Postgraduate Affiliate at the Vector Institute. I earned my B.Sc. in Mechanical Engineering from Sharif University of Technology and teach AI and machine learning at the University of Toronto and York University.
Non-provisional
Also contributed minor features and bug fixes to NVIDIA Warp.