Xihui Wu
Software Engineer at Google
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolXihui Wu is a software engineer with six years of experience building large-scale, production-ready systems across cloud, infrastructure, and ML domains at Google, Microsoft, AWS, and Amazon. He has deep hands-on experience in data engineering, ETL and serverless job execution, recommendation systems, and latency-sensitive backend services, complemented by contributions to Swift for TensorFlow where he implemented tensor initializers and a per-weight Adam optimizer. Based in Mountain View, he blends strong engineering fundamentals from a CS MS at The University of Chicago with practical product-focused delivery at major tech firms. Notably, his open-source work shows a penchant for low-level numerical and device-consistency fixes, indicating comfort at the intersection of systems and ML.
6 years of coding experience
5 years of employment as a software developer
Bachelor of Science (BS), Computer Science, Bachelor of Science (BS), Computer Science at Sun Yat-Sen UniversityMaster of Science (MS), Computer Science, Master of Science (MS), Computer Science at The University of Chicago
Github Skills (11)
swift10
machine-learning10
differentiable-programming10
gpt10
deep-learning10
tensorflow10
data-science9
optimizers9
optimizer9
xla9
datasets8
Programming languages (5)
Github contributions (5)
tensorflow/swift-models
Dec 2019 - Dec 2020
Models and examples built with Swift for TensorFlow
Role in this project:
ML Engineer
ML EngineerContributions:59 reviews, 46 commits, 90 PRs in 1 year
Contributions summary:Xihui primarily focused on refactoring and improving various machine learning models and related training pipelines within the Swift for TensorFlow framework. Their contributions included removing raw data access in multiple files, updating the usage of tensor initializers, and enabling source URL input for dataset creation. Moreover, the user made significant adjustments for training GPT2 model, and switching off old datasets.
swifttensorflowswift-for-tensorflow
tensorflow/swift-apis
Dec 2019 - Oct 2020
Swift for TensorFlow Deep Learning Library
Role in this project:ML Engineer
ML EngineerContributions:5 reviews, 13 commits, 15 PRs in 10 months
Contributions summary:Xihui contributed to the Swift for TensorFlow library by implementing and refining initialization methods for tensors. Their work included adding a categorical initializer and constraining it to TensorFlowFloatingPoint types. They also addressed issues related to the consistent device usage for bandpart operations within the library. Furthermore, the user added a per-weight Adam optimizer to the XLA optimizers.
differentiable-programmingswift-for-tensorflowdeep-learningmachine-learningdeep-learning-library
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.