Weiyi Wang
Staff Software Engineer at Google
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolWeiyi Wang is a Staff Software Engineer in the San Francisco Bay Area with a decade of experience building on-device ML and compiler infrastructure for TensorFlow Lite and LiteRT. At Google he has driven GenAI and mobile inference efforts—shipping runtime, converter, and compiler improvements for Gemini Nano and on-device generative AI. His contributions to the high-profile tensorflow/tensorflow repo include async APIs, metrics, and low-bit quantization (INT4 and I16 indices), reflecting deep expertise in performance-sensitive ML engineering. Prior roles span production systems at Facebook and infrastructure at Morgan Stanley, plus academic research and a conference publication on image retargeting. He combines systems-level thinking with practical ML model optimization, often working at the intersection of compiler, runtime, and device constraints.
10 years of coding experience
10 years of employment as a software developer
Bachelor of Science (BS) Computer Science, Bachelor of Science (BS) Computer Science at Fudan UniversityShanghai Southwest Weiyu Middle SchoolMaster of Science (M.S.) Software Engineering, Master of Science (M.S.) Software Engineering at Carnegie Mellon University
English, Shanghainese, Japanese, Chinese
Github Skills (14)
mle10
tensorflow-lite10
c-language10
cprogramming-language10
ml10
apidoc9
machine-learning9
api9
quantization9
deep-learning8
capi8
asynchronous8
model-optimization8
async8
Programming languages (2)
Github contributions (5)
tensorflow/tensorflow
Oct 2021 - Jan 2023
An Open Source Machine Learning Framework for Everyone
Role in this project:
Back-end Developer / ML Engineer
Back-end Developer / ML EngineerContributions:32 reviews, 83 commits, 30 comments in 1 year 3 months
Contributions summary:Weiyi primarily contributed to the TensorFlow Lite (TFLite) project, focusing on enhancing the framework's capabilities, particularly concerning asynchronous execution and quantization. They implemented modifications for the async API by adding accessor and updating task statuses. Additionally, the user added support for INT4 quantized weights and I16 indices, improving the performance of models within the TFLite environment. They also added various C API features for the asynchronous operation and metrics collection.
pythondata-sciencedeep-learningmlmachine-learning
cmusv-sc/SOC-Fall-2015-Team15-16-Lead-Fan-Luo-Mingyang-Ge
Oct 2015 - Dec 2015
Contributions:76 PRs, 90 pushes, 14 branches in 2 months
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.