Roshan Dathathri is a Principal Researcher at Microsoft with a decade of experience building optimizing compilers, runtimes, and distributed systems for heterogeneous and privacy-sensitive workloads. His work spans production-grade graph engines and memory-efficient graph representations to compilers for fully homomorphic encryption and encrypted neural network inference, published at top venues like PLDI and ASPLOS. He led engineering teams at Katana Graph to deliver low-latency, scalable distributed graph querying and a sparse collective communication layer, and now focuses on systems that make AI computation efficient across distributed, heterogeneous hardware. Known for bridging deep research with practical product impact, he combines formal compiler techniques with systems engineering to solve real-world performance and privacy challenges.
10 years of coding experience
9 years of employment as a software developer
Master of Science - MS, Computer Science, Master of Science - MS, Computer Science at Indian Institute of Science (IISc)
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at The University of Texas at Austin
Bachelor of Engineering - BE, Computer and Information Sciences, General, Bachelor of Engineering - BE, Computer and Information Sciences, General at Visvesvaraya Technological University
Galois: C++ library for multi-core and multi-node parallelization
Contributions:6 reviews, 938 commits, 28 PRs in 5 years 9 months
cudamultithreadingcppmulti-nodec-library
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.