Richard Wei
Manager, AI Frameworks (Foundation Models API)
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRichard Wei is an engineering leader with 11 years of experience building AI and language tooling, currently managing Apple's Swift AI Frameworks teams responsible for Foundation Models and Vision frameworks. He combines hands-on API design and cross-functional shipping experience—leading Foundation Models SDKs for Python and on-device Apple Intelligence features—with deep compiler and language expertise from contributions to Swift, SwiftSyntax, and Swift for TensorFlow. His background spans research-grade autodiff work at Google Brain to production ML framework and developer API launches at Apple, reflecting a rare blend of systems, ML, and compiler skill. An active open-source contributor, he has made notable contributions to prominent projects like Swift for TensorFlow and the Swift language itself, including macro integration and differentiable programming features. Colocated in San Francisco, he pairs rigorous technical authorship and documentation work with pragmatic product delivery across Apple platforms.
11 years of coding experience
8 years of employment as a software developer
University of Illinois Urbana-Champaign
English
Github Skills (33)
develop10
type-check10
optimizers10
abstract-syntax-tree10
macros10
compiler-plugin10
machine-learning10
optimizer10
compiler-design10
deep-learning10
tensorflow10
parsing10
computer-engineering10
parse10
mnist10
Programming languages (15)
Github contributions (5)
tensorflow/swift-apis
Feb 2019 - Sep 2019
Swift for TensorFlow Deep Learning Library
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 91 commits, 227 PRs in 7 months
Contributions summary:Richard primarily contributed to the Swift for TensorFlow Deep Learning Library by modifying code related to optimizers and layer implementations. They removed an MNIST test and added workarounds for specific issues related to the TensorFlow framework. Furthermore, the user simplified generic constraints and moved random number APIs and distributions into the DeepLearning module. Their changes also included performance measurement blocks and the modification of layer initializers.
differentiable-programmingswift-for-tensorflowdeep-learningmachine-learningdeep-learning-library
tensorflow/swift-models
Apr 2018 - Oct 2019
Models and examples built with Swift for TensorFlow
Role in this project:ML Engineer
ML EngineerContributions:42 commits, 96 PRs, 98 pushes in 1 year 5 months
Contributions summary:Richard made several significant contributions to the `swift-models` repository, focusing on the implementation and improvement of machine-learning models. They updated and refactored an MNIST model, integrating the `ParameterAggregate` synthesis. Furthermore, the user addressed bugs and optimized the code, as seen in the autoencoder model and the adoption of the new random API. The user also updated models to use the deep learning library, demonstrating expertise in applying machine learning techniques within the Swift for TensorFlow framework.
swifttensorflowswift-for-tensorflow
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.