Chun-an Chan
Software Engineer at Google DeepMind
Mountain View, California, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolChun-an Chan is a software engineer with a strong research background and three years of industry experience, currently working at Google DeepMind and contributing to text-to-speech research at Google. He holds a Ph.D. in Communication Engineering and previously led unsupervised spoken-language discovery and spoken-term indexing during a postdoc at National Taiwan University. Chun-an is an active contributor to major open-source ML infrastructure—his TFJS WASM work added a wide set of numerical kernels and 3D ops, improving browser-side performance and stability for a widely used TensorFlow.js backend. He blends low-level systems and DSP expertise (crypto on embedded Chrome OS and audio self-test tooling) with machine learning deployment skills, enabling production-ready model inference on constrained platforms. Based in Mountain View with roots in Taiwan and the UK, he brings a rare mix of academic rigor and practical engineering that accelerates ML at the edge.
3 years of coding experience
1 year of employment as a software developer
Bachelor's degree Electrical Engineering, Bachelor's degree Electrical Engineering at National Taiwan University
English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (15)
tensorflowjs10
blazor-webassembly10
c-language10
webassembly10
machine-learning10
cprogramming-language10
matrix9
backend9
linear-algebra9
data-structure9
back-end-development9
mat9
algorithm9
data-structures9
algorithms9
Programming languages (5)
Github contributions (5)
tensorflow/tfjs
Jan 2023 - Jan 2023
A WebGL accelerated JavaScript library for training and deploying ML models.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:151 reviews, 11 commits, 108 PRs in 19 days
Contributions summary:Chun-an's primary contribution involved implementing and integrating new kernels, specifically for WASM back-end, which includes trigonometric functions (atan, acos, asin, sinh), matrix operations (diag, bincount, denseBincount, broadcastArgs), 3D pooling operations (AvgPool3D, MaxPool3D, AvgPool3DGrad, MaxPool3DGrad), and tensor processing (SearchSorted, EluGrad, Dilation2D, Dilation2DBackpropFilter, Dilation2DBackpropInput, LRN, LRNGrad, Log1p, MaxPoolWithArgmax, Expm1, Multinomial, Unique, Conv3DBackpropFilterV2, ResizeBilinearGrad, and ResizeNearestNeighborGrad), along with related utility functions and configurations. They also addressed and resolved testing issues, ensuring the stability and functionality of the new kernels. The user's work directly contributes to expanding the computational capabilities and improving the performance of the TFJS WASM backend.
deployingjavascript-librarygpu-accelerationml-modelswebgl
chunnienc/tfjs
Jan 2023 - Jun 2023
A WebGL accelerated JavaScript library for training and deploying ML models.
Contributions:2 PRs, 320 pushes, 105 branches in 5 months
deployingjavascript-libraryml-modelswebgljavascript
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.