Victor Perez
Software Engineer at Meta
Old Toronto, Ontario, Canada
Summary
🤩
Rockstar🎓
Top SchoolVictor Perez is a software engineer with nine years' experience specializing in compilers, programming languages, and GPU code generation. He has driven production-grade compiler infrastructure at Codeplay and now contributes to ML compiler work at Meta, with notable open-source contributions to LLVM—implementing GPU-to-LLVM conversion passes, kernel metadata support, and Intel split-barrier handling. With academic foundations in formal methods from Universidad Politécnica de Madrid and hands-on research building static analysis tools like CiaoPP, he blends rigorous theory with practical compiler engineering. Based in Toronto, he thrives at the intersection of bleeding-edge tooling and low-level systems, often tackling the less-visible plumbing that makes high-performance ML and GPU workflows possible.
9 years of coding experience
4 years of employment as a software developer
Formal Methods in Computer Science and Engineering Mathematics and Computer Science, Formal Methods in Computer Science and Engineering Mathematics and Computer Science at Universidad Politécnica de Madrid
Spanish, English, Chinese
Github Skills (9)
spirv10
c-language10
compiler-construction10
gpu10
cprogramming-language10
llvm10
mlr9
opencl9
code-generation9
Programming languages (8)
Github contributions (5)
llvm/llvm-project
Jan 2022 - Jan 2025
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:122 reviews, 39 PRs, 40 pushes in 3 years
Contributions summary:Victor primarily contributed to the LLVM project by implementing and enhancing the GPU to LLVM conversion pipeline. Their work involved defining and implementing passes to convert GPU operations to LLVM dialect for the SPIR-V backend. They added support for kernel metadata, and also introduced conversion of gpu.func to llvm.func operations. They also worked on adding support for Intel's split barrier operations, demonstrating a focus on compiler infrastructure and GPU-related code generation.
compilerstechnologiesclangsubmittoolchain
victor-eds/intel-xpu-backend-for-triton
Sep 2024 - Jan 2025
OpenAI Triton backend for Intel® GPUs
Contributions:2 PRs, 95 pushes, 53 branches in 4 months
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.