Jin Sun
Engineering Leader at Meta
Bellevue, Washington, United States
Summary
👤
Senior🎓
Top SchoolJin Sun is an engineering leader with ~15 years of experience specializing in AI and big data, currently leading a 20+ engineer team at Meta responsible for FBLearner, the ML training orchestration backbone used across Ads, Recommendations, and Generative AI. He combines deep systems expertise from roles at Microsoft and Alibaba with hands-on contributions to open-source projects like pytorch/elastic, where he improved training reliability through robust synchronization and logging. Jin focuses on developer experience and operational excellence, driving iteration-speed gains that translated into hundreds of millions of dollars in value and accelerated Meta’s AI innovation. Known for scaling teams and systems—growing his group from 5 to ~20—he balances strategic leadership with low-level debugging and infrastructure design. Based in Bellevue, WA, he brings a proven track record of pushing technological boundaries in distributed ML and big data platforms.
10 years of coding experience
9 years of employment as a software developer
Master's degree Computer Science, Master's degree Computer Science at Tongji University
Github Skills (7)
pytorch10
python10
gloo10
distributed-computing10
pytest9
testing9
devops8
Programming languages (2)
Github contributions (5)
pytorch/elastic
Dec 2019 - Jan 2020
PyTorch elastic training
Role in this project:
Back-end Developer & DevOps Engineer
Back-end Developer & DevOps EngineerContributions:7 commits, 8 PRs in 1 month
Contributions summary:Jin primarily focused on improving the reliability and stability of the `pytorch/elastic` project, a PyTorch-based elastic training framework. Their contributions included addressing flaky tests by finding and using free ports to prevent conflicts. They also introduced a barrier mechanism using gloo to synchronize high-variance operations, which could resolve timeout or stuck issues in training. Furthermore, the user implemented event logging to improve debugging capabilities.
pytorchpythondeep-learningelasticmachine-learning
isunjin/flink
Sep 2018 - Apr 2019
Apache Flink
Contributions:46 pushes, 36 branches, 2 comments in 7 months
flinkapachebig-dataapache-flinkjava
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.