Learning & inference
Graph structure, statistical physics, information theory, and the geometry of learning.
Doctoral researchMachine learning, from first principles
to the physical world.
I work across machine learning, information theory, and scientific discovery. Currently, I’m building Venomics AI, exploring venom-derived molecules as a starting point for new therapeutics.
Interactive journey
Graph structure, statistical physics, information theory, and the geometry of learning.
Doctoral researchAdversarial robustness, dataset bias, and the gap between benchmark performance and the world.
Safety, ethics & alignmentVenom-derived therapeutics, frugal XR hardware, and embodied educational avatars.
Projects & experimentsCarnegie Mellon University. Graph structure as a prior for classification and detection.
I began my research career at IIT Madras, working on wireless communication, sparse channel estimation, and time-of-arrival estimation. My doctoral research at Carnegie Mellon University brought together spectral graph theory, statistical physics, network science, and information theory.
After work on medical time series and drug interactions, I became the first employee and Chief Scientist at UnifyID AI Labs. We studied human kinematics through motion-sensor data and ran five AI fellowship cohorts between 2016 and 2021.
At HAL51, which I co-founded with Sanghyun Han, I built frugal XR hardware and explored the practical problems of putting AI avatars in classrooms. These days, my attention is on biology.
Research, collaborations, or a good question.