Tianjun Xiao
Principal Machine Learning Engineer at NVIDIA
Shanghai, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolTianjun Xiao is a Principal Machine Learning Engineer with 11 years of experience building large-scale ML and computer vision systems for leading tech teams, now at NVIDIA after senior roles at AWS Shanghai AI Lab, Tesla Autopilot, and Microsoft. He combines deep research roots from MSRA and a Peking University CS master's with hands-on engineering—implementing and optimizing deep learning models, data ingestion pipelines, and production-grade ML infrastructure. His open-source work spans influential projects like MXNet, DGL, and the multi-GPU Minerva framework, where he contributed image-classification model implementations and clarified API documentation across frameworks. Tianjun excels at bridging research and product: translating novel model ideas into robust, deployable pipelines for real-world perception and recommendation systems. Colleagues know him for pragmatic architecture choices and an ability to improve developer experience through clear docs and tooling. He is based in Shanghai and brings a rare mix of low-level framework experience and production ML leadership.
11 years of coding experience
5 years of employment as a software developer
High School Affiliated to Nanjing Normal UniversityBachelor's degree Computer Science, Bachelor's degree Computer Science at Nankai UniversityMaster's degree Computer Science, Master's degree Computer Science at Peking University
Chinese
Github Skills (14)
neural-network10
computer-vision10
mxnet10
machine-learning10
api-documentation10
convolutional-neural-networks10
deep-learning10
graph-neural-network10
python10
image-processing10
imagenet10
documentation10
tensorflow7
pytorch3
Programming languages (3)
Github contributions (5)
dmlc/minerva
Nov 2014 - Aug 2015
Minerva: a fast and flexible tool for deep learning on multi-GPU. It provides ndarray programming interface, just like Numpy. Python bindings and C++ bindings are both available. The resulting code can be run on CPU or GPU. Multi-GPU support is very easy.
Role in this project:
ML Engineer
ML EngineerContributions:112 commits, 1 PR, 59 pushes in 8 months
Contributions summary:Tianjun's commits primarily involve modifications and additions to an image classification application within the Minerva deep learning framework. They focus on adapting the application for the ImageNet dataset, suggesting an interest in computer vision tasks. The code changes reveal work on defining and training a model (AlexNet), including its architecture, forward/backward passes, and loss calculations. These contributions suggest a focus on implementing and optimizing deep learning models using the Minerva framework.
pythonndarraygpu-supportpython-numpymulti-gpu
apache/mxnet
Aug 2015 - Nov 2015
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:Back-end Developer
Back-end DeveloperContributions:110 commits, 9 PRs, 13 pushes in 2 months
Contributions summary:Tianjun's commits focus on implementing features within the image recordio iterator. They implemented a MNIST iterator and created a TBlobBatch to handle the data. These commits demonstrate the user's work on data loading and processing pipelines, specifically related to the core data ingestion aspects of a deep learning project.
pythonschedulerdataflowmutationdata-science
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.