Selva Chandrasekaran Selvaraj

About

Selva Chandrasekaran Selvaraj is a computational materials and molecular scientist with 8+ years of experience building GPU-accelerated AI/ML software for energy materials and molecular design. His research combines quantum chemical simulations and machine learning to address challenging problems in solid-state battery electrolytes, semiconductor defects, and catalytic surfaces, deploying large-scale Machine Learning Molecular Dynamics (MLMD) at 100× the speed of conventional DFT on H100 GPU clusters at NREL and Argonne National Laboratory. 

Selva's work spans the full AI materials pipeline: from training ML interatomic potentials (DeepMD-kit, MACE-MPA-0, AIMNet2) and running nanosecond-scale MLMD on systems of one million atoms, to graph neural networks for multi-property crystal prediction and generative AI (graph diffusion DDPM) for novel cathode design.