Summary
Aditya Bhagwat is a software engineer specializing in systems ML and compiler work, currently accelerating deep learning on Meta’s custom AI accelerator. With nine years of experience across startups and research—Modular, Snap, and CMU—he blends compiler engineering (MLIR), distributed systems, and large-scale PyTorch training to optimize performance for specialized hardware. He has hands-on experience in reducing communication overhead in distributed training and building infrastructure to train graph models at massive scale, demonstrating both research rigor and production focus. Comfortable in C++ and parallelism, he pairs an interest in profiling with practical library and runtime development. Based in Pittsburgh, he brings a rare combination of compiler-level expertise and applied ML systems know-how that speeds model training from research prototypes to chip-optimized deployments.
10 years of coding experience