Bowen Bao
Principal MTS Software Engineer at AMD
Greater Seattle Area United States
Summary
🤩
Rockstar🎓
Top SchoolBowen Bao is a Principal MTS Software Engineer with 11 years of experience building and optimizing AI frameworks and platform tooling, now leading AI Frameworks at AMD after a multi-year tenure on PyTorch, ONNX and ONNX Runtime at Microsoft. He specializes in backend systems for ML inferencing and model interchange—contributing to high-profile open-source projects like ONNX Runtime and PyTorch where he improved exporter stability, shape inference, and diagnostics. Bowen’s work spans both deep algorithmic fixes (operator support, gradient corrections) and developer experience improvements (testing utilities, logging and stack-trace diagnostics), helping bridge research models to production runtimes. He holds an M.S. in Computer and Information Science from the University of Pennsylvania and augmented his training with a visiting term at Columbia, reflecting strong academic foundations. Colleagues value his blend of meticulous bug-fixing and pragmatic system design that surfaces subtle correctness issues across complex operator stacks. Less obvious: he often focuses on the plumbing—type promotion, shape inference, and test utilities—that prevents hard-to-reproduce failures in production ML pipelines.
11 years of coding experience
6 years of employment as a software developer
Bachelor of Engineering, Computer Software Engineering, 3.35, Bachelor of Engineering, Computer Software Engineering, 3.35 at East China Normal UniversityVisiting Student, Computer Science, 4.0, Visiting Student, Computer Science, 4.0 at Columbia University in the City of New YorkMaster’s Degree, Computer and Information Science, 4.0, Master’s Degree, Computer and Information Science, 4.0 at University of Pennsylvania
Github Skills (19)
debugging10
pytorch10
debug10
c-language10
python10
machine-learning10
cntk10
inference10
onnx10
deep-learning10
tensorflow10
cprogramming-language10
shapes10
testing9
code-optimization9
Programming languages (6)
Github contributions (5)
onnx/onnx
Sep 2018 - Mar 2021
Open standard for machine learning interoperability
Role in this project:
ML Engineer
ML EngineerContributions:17 reviews, 21 commits, 29 PRs in 2 years 6 months
Contributions summary:Bowen primarily contributed to the ONNX repository by modifying shape inference logic and adding support for new operators and features related to machine learning model interoperability. Their work focused on enhancing the functionality of existing operators like `Clip`, `Reshape`, and `Det`, adding new operators like `ScatterND`, and updating the shape inference tests. These changes facilitate the conversion and optimization of machine learning models for diverse hardware and software platforms.
pytorchmxnetdeep-learninginteroperabilitymachine-learning
pytorch/pytorch
Oct 2018 - Jan 2023
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:2130 reviews, 940 commits, 900 PRs in 4 years 4 months
Contributions summary:Bowen primarily contributed to the development and refactoring of the ONNX exporter, focusing on improving its functionality and stability for PyTorch models. They implemented features to support the export of custom operators and improved handling of various data types. A significant part of their work involved the development of diagnostics tools, including enhancements to logging and the incorporation of stack trace information, aiding in debugging and model validation. They also worked on integrating a type promotion system, ensuring the compatibility of different data types during export.
pythongpu-accelerationdeep-learninggpunumpy
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.