Yao Shi
Senior Software Engineer at NVIDIA
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolYao Shi is a Senior Software Engineer and architect with nine years of industry experience, currently advancing deep learning compilers at NVIDIA from his base in Seattle. He holds a PhD in Computer Science from Tsinghua University and combines strong research roots in program analysis and worst-case execution time with practical system-building across compilers, VPUs, and GPU codegen. At Intel he was a core contributor to the open-source PlaidML project, implementing IR-level performance optimizations and a constraint-reduction pass that reduced redundant constraints and improved codegen tiling. His background spans both academia and industry—from microkernel research and automatic bug detection to leading engineering teams and advising startups—giving him a rare blend of rigorous analysis and product-focused delivery. Comfortable moving between reinforcement learning algorithms, ML systems, and low-level performance tuning, he often applies ideas from arts and design to craft elegant engineering solutions.
10 years of coding experience
12 years of employment as a software developer
Bachelor's degree, Computer Science, Bachelor's degree, Computer Science at Tsinghua University
Chinese, English
Github Skills (10)
algorithm10
data-structures10
algorithms10
tiles10
c-language10
cprogramming-language10
plaidml10
data-structure10
machine-learning9
deep-learning9
Programming languages (2)
Github contributions (5)
plaidml/plaidml
Mar 2019 - May 2022
PlaidML is a framework for making deep learning work everywhere.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:46 reviews, 194 commits, 110 PRs in 3 years 2 months
Contributions summary:Yao contributed to the PlaidML repository by adding a constraint reduction pass and other supporting features. They implemented a new pass that can remove redundant constraints and reduce the index range and also added the class to support alias for accessing data. The user's code additions included improvements for the tile/codegen package.
deep-learningplaidml
SwxllowAG/plaidml
Sep 2019 - Oct 2019
PlaidML is a framework for making deep learning work everywhere.
Contributions:3 commits in 1 month
deep-learningplaidml
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.