Yan Facai
Senior Algorithm Engineer at ByteDance
Haidian District, Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolYan Facai is a Senior Algorithm Engineer with 11 years of experience building and optimizing large-scale ML systems, currently leading algorithm work at ByteDance in Beijing. A committed open-source contributor, he is a Google TensorFlow member and has added layers and metric-learning losses to tensorflow/addons while improving MLlib algorithms in Apache Spark. His hands-on contributions to Angel-ML’s parameter server and logistic regression optimizations show deep expertise in distributed training and gradient-based optimization. Previously at Alibaba and Weibo, he focused on productionizing recommendation and search algorithms, bridging research and engineering. Trained in Instrument Science and Engineering at Shanghai Jiao Tong University, he combines rigorous measurement-background thinking with practical system-level bug fixes and performance tuning. Notably, his work often targets subtle correctness and usability issues—intercept handling, model loading, and impurity calculations—that materially improve model reliability in production.
11 years of coding experience
3 years of employment as a software developer
Bachelor’s Degree Measuring and Control Technology and Instrumentations, Bachelor’s Degree Measuring and Control Technology and Instrumentations at Shandong UniversityMaster’s Degree Instrument Science and Engineering, Master’s Degree Instrument Science and Engineering at Shanghai Jiao Tong University
Chinese, English, Japanese
Stackoverflow
51reputation
3kreached
3answers
0questions
Github Skills (22)
spark10
gradient-descent10
python10
big-data10
machine-learning10
ml10
scala10
trainings10
tensorflow10
neural-network10
params10
decision-tree10
logistic-regression10
para10
random-forest10
Programming languages (10)
Github contributions (5)
tensorflow/addons
Jan 2019 - Oct 2019
Role in this project:
ML Engineer
ML EngineerContributions:3 releases, 43 commits, 100 PRs in 9 months
Contributions summary:Yan's commits primarily focused on enhancing the `tensorflow/addons` repository with new functionality related to machine learning. They implemented a Maxout layer and a PoincareNormalize layer, suggesting a focus on model architecture and feature engineering. Furthermore, the user integrated the triplet semi-hard loss, indicating contributions to loss function development within the context of metric learning.
python2-xdeep-learningaddonssig
apache/spark
Jul 2017 - Aug 2017
Apache Spark - A unified analytics engine for large-scale data processing
Role in this project:Data Scientist
Data ScientistContributions:3 commits, 21 PRs, 137 comments in 11 days
Contributions summary:Yan contributed to the Apache Spark project by implementing and refining machine learning algorithms. They focused on bug fixes and enhancements within the `MLlib` module, specifically improving the DecisionTreeModel and RandomForest implementations. These contributions involved addressing issues with impurity calculations, split value handling, and parameter settings to enhance the functionality and accuracy of the machine learning models. Additionally, the user exposed a parameter in PySpark FPGrowth.
analyticspythondata-processingsqlapache
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.