Zhenhua Han
Scientist at Nex-AGI
Shanghai, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolZhenhua Han is a scientist and senior researcher with 17 years of experience building systems for machine learning, currently working at Nex-AGI after multi-year roles at Microsoft Research Asia. He holds a PhD in Computer Science from The University of Hong Kong and has a background spanning electrical engineering and applied optimization for cloud and wireless systems. At Microsoft he contributed to production-grade AutoML tooling—notably optimizations to the widely used open-source NNI toolkit—improving multi-model training, device placement, and cross-graph deduplication. His work blends deep research rigor with practical engineering, addressing CUDA/PyTorch compatibility and trainer validation to make ML workflows more efficient and robust. Based in Shanghai, he combines academic depth with hands-on system design, often surfacing subtle infrastructure fixes that unlock large performance gains.
17 years of coding experience
5 years of employment as a software developer
Bachelor of Science (BS) Electrical and Electronics Engineering, Bachelor of Science (BS) Electrical and Electronics Engineering at City University of Hong KongThe University of Hong Kong (HKU)Bachelor of Science (BS) Electrical and Electronics Engineering, Bachelor of Science (BS) Electrical and Electronics Engineering at University of Electronic Science and Technology of China
English, elementary japanese
Github Skills (11)
pytorch10
machine-learning10
deeplearning-ai10
deep-learning10
pytorch-lightning10
automated-machine-learning10
python10
neural-architecture-search10
model-compression9
data-science8
tensorflow8
Programming languages (6)
Github contributions (5)
microsoft/nni
Sep 2020 - May 2022
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:
ML Engineer
ML EngineerContributions:46 reviews, 27 commits, 12 PRs in 1 year 8 months
Contributions summary:Zhenhua's contributions primarily involve developing and optimizing the NNI (Neural Network Intelligence) toolkit, which automates the machine learning lifecycle. They have focused on optimizing the Retiarii module, improving device placement strategies, implementing input deduplication for cross-graph optimization, and integrating validation in the base trainers. These enhancements aimed to enhance performance and efficiency, specifically for multi-model training scenarios. The user also addressed bugs related to CUDA ordinal handling and compatibility across different versions of PyTorch-Lightning, demonstrating a strong understanding of the framework.
pythonneural-architecture-searchengineeringtensorflowbayesian-optimization
hzhua/spark-monotasks
Nov 2017 - Dec 2020
Fast, predictable data analytics based on (and API-compatible with) Apache Spark
Contributions:2 PRs, 1 push, 1 branch in 3 years 1 month
apianalyticsdata-analyticsdata-ingestionapache
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.