Summary
Kumudha Narasimhan is a technical and people leader with 10 years of experience optimizing AI and HPC performance across diverse architectures, currently leading compiler work at Riverlane from Cambridge. She has driven cross-platform SYCL and oneDNN/oneMKL integrations, led multi-team efforts to port GPU backends and optimize model runtimes, and published peer-reviewed research on DNN portability and tile-size selection. At Codeplay she managed 18+ engineers across specialized Agile teams, coordinated division-level OKRs and EU projects, and improved engineering practices for reproducibility and ISO readiness. Kumudha blends hands-on compiler and performance engineering with mentorship and organizational design, creating inclusive teams that deliver on strategic goals. She is comfortable presenting technical results at conferences and translating research into customer-facing optimizations, a trait that helped accelerate SYCL adoption in production stacks. Her background—MSc in High Performance Computing and Compilers from IISc and early experience in finance technology—gives her a rare mix of rigorous research, production engineering, and cross-domain collaboration.
10 years of coding experience
7 years of employment as a software developer
Master of Science (Engineering) High Performance Computing and Compilers, Master of Science (Engineering) High Performance Computing and Compilers at Indian Institute of Science (IISc)
Bachelor’s Degree Computer Science and Engineering, Bachelor’s Degree Computer Science and Engineering at M S Ramaiah Institute of Technology
English, Hindi, Tamil, Kannada