Zifei Shan
Engineering Manager, Expert Applied Scientist (T12) at Tencent
Shanghai, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolZifei Shan is an engineering leader and applied scientist with 12 years of experience building ML systems that bridge research and production at scale, currently leading LLM landing and the WXG LLM platform at Tencent. He has a track record of turning research into product across Apple, Google Research, and the startup Lattice (acquired by Apple), focusing on knowledge graphs, information extraction, and recommender user-interest representations. Trained at Stanford (MS) and Peking University (BS), he was an active contributor to DeepDive, helping scale probabilistic inference and in-database extractors for real-world data pipelines. Zifei combines hands-on back-end contributions and platform leadership—having led developer productivity and post-training/SFT/RL/distillation efforts—to accelerate deployment of high-quality ML at internet scale. Colleagues describe him as equally at home in code, research, and large-team strategy, with a knack for turning “dark data” into structured, actionable knowledge. He’s based in Shanghai and continues to hire and grow engineering teams across China.
12 years of coding experience
10 years of employment as a software developer
Research Exchange Student, Electrical and Electronics Engineering, Research Exchange Student, Electrical and Electronics Engineering at Technion - Israel Institute of TechnologyBachelor of Science (BS), Computer Science, Bachelor of Science (BS), Computer Science at Peking UniversityMaster of Science (MS), Computer Science, Master of Science (MS), Computer Science at Stanford University
Chinese, English, Japanese
Github Skills (4)
postgresql10
python8
mysql8
database-design7
Programming languages (3)
Github contributions (5)
HazyResearch/deepdive
Jan 2014 - Jan 2017
DeepDive
Role in this project:
Back-end Developer
Back-end DeveloperContributions:2 releases, 412 commits, 39 PRs in 3 years
Contributions summary:Zifei primarily worked on modifying and enhancing a tutorial for the DeepDive project. They implemented code changes to support various aspects, including adding features related to skipping learning and setting initial weights, and supporting faster in-database extractors. The modifications touched upon key components of the system, indicating contributions to back-end functionality and data processing.
deep-learning
zifeishan/cs224s-deepSpeech
Apr 2014 - Jun 2014
Contributions:42 commits in 1 month
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.