Billy Zhu is a compiler engineer with a decade of experience building production-quality tooling for AI and systems software, currently working on heterogeneous compiler and runtime integration at Modular. He previously led compiler development for SambaNova’s reconfigurable dataflow hardware and has deep hands-on experience in MLIR/LLVM, including work on debug-info import, caching, and leak fixes that improved compiler performance and reliability. Billy also has full-stack roots from supporting large-scale autograding infrastructure at Carnegie Mellon’s Autolab, where he shipped backend, API, and UX fixes used by thousands of students. Comfortable across C++, Python, and web stacks, he brings a pragmatic balance of systems-level rigor and product-minded engineering; unusually, his background spans both low-level compiler internals and user-facing education tooling.
10 years of coding experience
7 years of employment as a software developer
Master of Science - MS, Computer Science, Master of Science - MS, Computer Science at Carnegie Mellon University
High School, High School at Beijing National Day School
Course management service that enables auto-graded programming assignments.
Role in this project:
Full-stack Developer
Contributions:35 commits, 78 PRs, 124 pushes in 2 years 7 months
Contributions summary:Billy made several contributions focused on improving the Autolab course management system. They addressed a trailing slash issue in the assessment import process and implemented a caching mechanism for the gradesheet page to reduce server load. The user also made UI/UX improvements by adding full name search capabilities and fixing date picker variables. They also addressed minor bugs by removing redundant calls and updating gradesheet sorting.
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
Role in this project:
Back-end Developer
Contributions:62 reviews, 46 PRs, 27 pushes in 1 year 3 months
Contributions summary:Billy contributed to the LLVM project, focusing on the MLIR subproject. Their work involved implementing and refining debug information import features, specifically addressing recursive dependencies and caching mechanisms to improve performance. They added a new debug name table kind to the DICompileUnit and addressed memory leaks in the export cache. The user also fixed issues related to call site locations and compiler optimizations.
compilerstechnologiesclangsubmittoolchain
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.