Harsha Lokavarapu is a software engineer with 11 years of experience bridging high-performance scientific computing and practical machine learning at scale. He has deep expertise in CUDA-accelerated numerical methods and distributed HPC workflows, and translated that background into training complex TensorFlow/Keras models for imbalanced multi-label image tasks and Kaggle competitions. At Uber he built Go services and tooling to emulate data-center network topologies and managed lab infrastructure, and he now contributes to engineering at Motive. His work uniquely combines production systems engineering with research-grade experiments on supercomputers and multi-GPU cloud setups, including published results in geophysics. Comfortable across C++, CUDA, Go, and Python, he mentors junior researchers and engineers and leverages analytic rigor (a UC Davis MS in Geophysics and BS in CS) to improve model performance and computational scalability. An understated strength is his ability to diagnose low-level bottlenecks—whether particle generation in PIC codes or GPU training inefficiencies—and turn them into repeatable solutions.
11 years of coding experience
4 years of employment as a software developer
Minor, Applied Mathematics, Minor, Applied Mathematics at University of California, Davis
Contributions:1 PR, 146 pushes, 75 branches in 5 years 11 months
physics3d-modelsthermalparallelproblems
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.