Xueyun Zhu
Software Engineer at Apple
Redmond, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolXueyun Zhu is a software engineer with seven years of experience building high-performance machine learning infrastructure and runtime systems, currently based in Redmond and working at Apple. Previously at Microsoft, Xueyun contributed to Azure AI Platform projects like ONNX Runtime and ONNX.js, focusing on WebGL GPU optimizations, shader work, and improving build and training integration. Their open-source contributions include adding texture packing/unpacking and kernel optimizations to microsoft/onnxjs and resolving complex build and merge issues in microsoft/onnxruntime, reflecting deep backend and MLOps expertise. Trained in electrical and computer engineering at Georgia Tech and Peking University, they blend low-level systems understanding with practical ML deployment skills. Colleagues can expect a pragmatic engineer who dives into shaders, performance bugs, and CI/build fixes that unlock real-world model acceleration.
7 years of coding experience
8 years of employment as a software developer
B.S, Electronics, Electrical and Computer Engineering, B.S, Electronics, Electrical and Computer Engineering at Peking University (PKU)M.S, Electrical and Computer Engineering, M.S, Electrical and Computer Engineering at Georgia Institute of TechnologyB.S., Electronics, B.S., Electronics at Peking University
Chinese
Github Skills (21)
webgl10
javascript10
c-language10
shader10
machine-learning10
onnx10
glsl10
tensorflow10
gpu10
optimisation10
graphics-programming10
cprogramming-language10
optimization10
build-system9
tensorrt9
Programming languages (5)
Github contributions (5)
microsoft/onnxjs
Jan 2021 - Apr 2021
ONNX.js: run ONNX models using JavaScript
Role in this project:
Back-end Developer
Back-end DeveloperContributions:67 reviews, 109 commits, 19 PRs in 3 months
Contributions summary:Xueyun focused on enhancing the WebGL backend for the ONNX.js library. Their work primarily involved adding support for texture packing and unpacking operations, crucial for optimizing model execution on GPUs. They implemented new shader code and coordinate libraries, addressing bugs and improving the functionality of existing kernels. Furthermore, the user integrated these pack/unpack features into binary and unary operations to optimize WebGL performance.
javascriptonnx-modelsonnx
microsoft/onnxruntime
Mar 2020 - Jun 2021
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:Back-end Developer & MLOps Engineer
Back-end Developer & MLOps EngineerContributions:1 release, 86 reviews, 67 commits in 1 year 3 months
Contributions summary:Xueyun primarily worked on resolving merge conflicts and fixing build breaks within the ONNX Runtime project. Their contributions involved modifications to core files, including those related to tree ensemble aggregation, and training-related components. The commits indicate a focus on improving build processes, and integration of model training and inference. The code changes also suggest involvement in testing and optimizing the build process.
runtimetrainingtensorflowai-frameworkaccelerator
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.