Summary
Patrick Johnstone is a Machine Learning Engineer with a strong academic foundation (PhD) and 8+ years of applied research experience translating advanced optimization theory into high-performance ML systems. He has a proven track record in convex and nonconvex algorithm design, with seven first‑author journal publications and over 100 citations, and has driven distributed optimization work on exascale HPCs at Brookhaven. At Meta he applies this expertise to Ads Supply ML and Foundations, combining hands-on PyTorch and MPI/Slurm engineering with formal verification skills using the Lean theorem prover to improve algorithm reliability. His background includes industry research internships (Qualcomm, Rambus) and award-winning practical work, and he has taught graduate and undergraduate ML and signal processing courses, bridging rigorous theory with production-ready implementations.
8 years of coding experience
6 years of employment as a software developer
Bachelor's Degree, Electrical, Electronics and Communications Engineering, University Medal, Bachelor's Degree, Electrical, Electronics and Communications Engineering, University Medal at University of New South Wales
University of Illinois Urbana-Champaign