Pankaj Rajak is an Applied Scientist II with 11 years of experience specializing in reinforcement learning, generative models, and geometric neural networks for inverse design of materials with target physical properties. His work spans industry and national lab settings—building offline RL agents for autonomous material synthesis at Argonne and advancing graph neural network force fields and representation learning for molecular simulation. During a PhD at USC he scaled reactive molecular dynamics to billions of atoms, developed force fields via multi-objective optimization, and applied VAEs and normalizing flows to 3D point-clouds for materials discovery. At Amazon he bridges cutting-edge research with production ML, bringing a rare combination of deep physics-based simulation expertise and modern AI methods. He also holds an AI graduate certificate from Stanford and blends meta-learning insights (e.g., task-memorization issues in few-shot learning) into practical workflows for accelerated materials discovery.
11 years of coding experience
9 years of employment as a software developer
Phd, Computational Materials Science, Phd, Computational Materials Science at University of Southern California
Dual Degree: B.tec(Hons)+M.tec, Metallurgical and Materials Engineering, Dual Degree: B.tec(Hons)+M.tec, Metallurgical and Materials Engineering at Indian Institute of Technology, Kharagpur
Artificial Intelligence Graduate Certificate, 4.3, Artificial Intelligence Graduate Certificate, 4.3 at Stanford University
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.