Yu Deshui
Senior Tech Expert Team Lead at Alibaba Group
Chaoyang District, Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolYu Deshui is a Senior Tech Expert and team lead with 8+ years building high-availability backend systems and leading engineering teams at Alibaba after formative roles at Amazon, Microsoft, and IGT. He architected and ran Xiaoyun, an intelligent assistant backend reaching 10M DAU with a 99.9% SLA, and led a 10-engineer team to deliver a high-throughput online advertising system in three months. Technically versatile, he contributes to open-source AutoML tooling (notably enhancements to microsoft/nni’s PAI training service and DevOps automation) and focuses on backend, DevOps, and integration with speech and NLU services. His background combines high-performance computing from Tsinghua with hands-on product delivery across cloud, mobile, and distributed systems. Known for driving rapid, reliable releases and pragmatic infrastructure design, he blends deep technical ownership with team leadership to convert research-grade components into production-scale services.
8 years of coding experience
7 years of employment as a software developer
和君商学院7届Master, Computer Science and Technology, High Performance Computing Institute, Master, Computer Science and Technology, High Performance Computing Institute at Tsinghua University
English
Github Skills (13)
hdfs10
python10
dockers9
automl9
cicd9
automated-machine-learning9
distributed-training9
docker9
typescripts8
typescript8
typescript-types8
kubernetes-pods7
kubernetes7
Programming languages (5)
Github contributions (5)
microsoft/nni
Aug 2018 - Jun 2019
An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.
Role in this project:
Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:6 releases, 63 commits, 160 PRs in 9 months
Contributions summary:Yu's commits primarily involve code changes to the NNI manager, focusing on implementing and enhancing the PAI training service. They implemented the PAI training service, added a trial-keeper module, and made significant updates to the PAI integration, including HDFS and token management. Further contributions include bug fixes, enhancements to the PAI scripts and Dockerfiles, and the addition of installation scripts. They were also involved in improving file organization and upload efficiency related to code directories.
pythonneural-architecture-searchengineeringtensorflowbayesian-optimization
yds05/nni
Sep 2018 - Sep 2018
An open source AutoML toolkit for neural architecture search and hyper-parameter tuning.
Contributions:4 PRs, 7 pushes, 1 branch in 1 day
pythonneural-architecture-searchtensorflowhyperparameter-tuning
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.