Sijun Tan
PhD Candidate
Berkeley, California, United States
Summary
🤩
Rockstar🎓
Top SchoolSijun Tan is a CS PhD candidate at UC Berkeley’s Sky Computing Lab with eight years of experience building scalable, privacy-preserving systems and AI agents. He has industrial experience as a Senior Algorithm Engineer at Ant Group working on MPC-based secure computation and contributed performance-critical backend optimizations to Facebook Research’s CrypTen, including CUDA and 2PC equality improvements. His research spans mechanism design, algorithmic game theory, computer vision, and generalist AI agent scaling, grounded in strong theoretical training (BS CS and BA Math, UVA, 3.95 GPA). Based in Berkeley, he blends research rigor with production engineering, shipping cryptographic and ML optimizations used in real systems. Outside typical academic roles, he’s notable for translating research papers directly into practical code optimizations that materially improve secure ML frameworks.
8 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of California, BerkeleyThe Affiliated High School of South China Normal UniversityBachelor of Science - BS, Computer Science, 3.95/4, Bachelor of Science - BS, Computer Science, 3.95/4 at University of Virginia
Chinese, English
Github Skills (9)
code-optimization10
pytorch10
machine-learning10
optmization10
python10
optimisation10
optimization10
cuda9
algorithms9
Programming languages (7)
Github contributions (5)
facebookresearch/CrypTen
May 2020 - Aug 2020
A framework for Privacy Preserving Machine Learning
Role in this project:
Back-end Developer / ML Engineer
Back-end Developer / ML EngineerContributions:6 reviews, 23 commits, 35 PRs in 2 months
Contributions summary:Sijun primarily contributed to optimizing the `crypten` framework, focusing on privacy-preserving machine learning. They optimized the equality function for the 2PC case, leveraging a research paper for improved performance. Further contributions included integrating more parameters into the MPCConfig and implementing CUDA integration and optimizations for the TFP component. These changes demonstrate a focus on enhancing the framework's functionality and efficiency for secure computation.
privacydeep-learningprivacy-preserving-machine-learningdifferential-privacymachine-learning
jeffreysijuntan/CrypTen
Dec 2019 - Jul 2020
A framework for Privacy Preserving Machine Learning
Contributions:2 PRs, 3 pushes, 20 branches in 6 months
privacy-enhancing-technologiesprivacyprivacy-preserving-machine-learningmachine-learningprivacy-preserving
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.