Tang, Cheng
Sunnyvale, California, United States
Summary
🤩
RockstarCheng Tang is a Partner Software Engineer at Microsoft with 11 years of experience building high-performance back-end systems and ML runtime components. He has progressed through senior engineering roles at Microsoft after earlier system-level work at MicroStrategy, combining production-grade engineering with platform architecture. Cheng is a frequent open-source contributor to prominent ML projects—his backend work on apache/tvm, keras, and onnxruntime improved cross-platform support, added bfloat16 ops, and hardened Windows and CNTK integrations. He specializes in compiler and runtime engineering for ML inference and training, with practical expertise in bridging OS-level details and high-level model semantics. Based in California and trained at Peking University, he brings a pragmatic systems mindset and a track record of shipping nuanced cross-platform fixes that prevent subtle runtime failures.
11 years of coding experience
Github Skills (19)
c-language10
back-end-development10
python10
machine-learning10
cntk10
onnx10
compiler-compiler10
deep-learning10
neural-network10
compiler10
backend10
cprogramming-language10
windows9
cuda8
deeplearning-ai8
Programming languages (5)
Github contributions (5)
microsoft/onnxruntime
Nov 2018 - Jan 2023
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:650 reviews, 104 commits, 171 PRs in 4 years 2 months
Contributions summary:Tang, primarily contributed to the ONNX Runtime library, focusing on machine learning model inferencing and training. Their work involved modifications to the core codebase, specifically addressing character encoding issues, integrating external libraries like BrainSlice, and updating kernel memory type interfaces. The user also added support for bfloat16 data type in ops and implemented improvements in the compilation process.
runtimetrainingtensorflowai-frameworkaccelerator
keras-team/keras
Jun 2017 - Apr 2018
Deep Learning for humans
Role in this project:Back-end Developer
Back-end DeveloperContributions:11 commits, 18 PRs, 84 comments in 10 months
Contributions summary:Tang, primarily contributed to the `keras-team/keras` repository by addressing issues related to the CNTK backend. Their commits focused on improving error messages, fixing incorrect behavior in the batch normalization layer, and enabling stateful RNNs. The user also worked on integrating the CNTK backend with the latest CNTK releases, including supporting new APIs and features.
pythondata-sciencedeep-learningneural-networksmachine-learning
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.