About
I am an applied machine learning researcher and engineer working at the intersection of 3D computer vision, geometric learning, generative modeling, and physical AI.
I currently work at Dandy, where I lead projects involving 3D point clouds and meshes, and bridge 3D compute vision methods with modern represetnation learning methods. Some of the things I’ve worked on are 3D generative modeling using conditional diffusion models, 3D shape retrieval using 3D scans and SE(3)-equivariant architectures, and 2D-to-3D registration.
Before this, I worked on large-scale machine learning and statistical modeling at Jain Global, Bloomberg, Two Sigma, and AT&T Labs. My background is in electrical and computer engineering, signal processing, probability, statistics, linear algebra and information theory. I hold a Ph.D. WINLAB from Rutgers University and a B.Tech. in Electrical Engineering from IIT Madras.
My current interests include:
- 3D vision and geometric deep learning
- Generative models for structured physical objects
- Inverse problems and uncertainty
- AI4Science (weather, health, biology)
- Reliable evaluation and deployment of ML systems
If you are interested in chatting about any of these areas, feel free to drop me a line at
My work has resulted in 8 U.S. patents (one licensed to the Ford Motor Company), best-paper awards at ACM MobiSys and NYU CSAW, and publications across wireless systems, network security, information theory, and machine learning.
This website is where I write about machine learning, geometry, sensing, biology, medicine, and scientific inference.