Eric Junyuan Xie
Software Engineer at University of Washington
Seattle, Washington, United States
Summary
🤩
Rockstar
Top expert inHigh-Performance Machine Learning Computing
Eric Junyuan Xie is a seasoned software engineer with 14 years of experience building high-performance back-end and ML systems, currently at Bytedance in Seattle. He has deep C++ and CUDA expertise, contributing to major open-source projects like MXNet, TVM and mshadow where he implemented autograd features, optimized GPU kernels, and extended multi-dimensional tensor support. Eric blends systems-level engineering with machine learning research—regularly improving compilers, memory planning, and distributed training infrastructure while also polishing educational notebooks for deep learning. He’s comfortable across the stack from CI/CD and build systems to low-level operator implementation, and often surfaces hard-to-find stability and performance fixes that make complex ML tooling production-ready.
14 years of coding experience
Github Skills (43)
c-language10
operation10
python10
tensorrt10
gpu-programming10
mxnet10
machine-learning10
deep-q-learning10
bash10
dm10
cicd10
kernel10
mask-rcnn10
deeplearning-ai10
compiler-compiler10
Programming languages (5)
Github contributions (5)
piiswrong/deep3d
Apr 2016 - Apr 2016
Automatic 2D-to-3D Video Conversion with CNNs
Role in this project:
ML Engineer
ML EngineerContributions:14 commits, 11 pushes, 1 branch in 11 days
Contributions summary:Eric primarily focused on updating and refining the `deep3d.ipynb` notebook, which involves automatic 2D-to-3D video conversion using CNNs. Their contributions included updates to the notebook's content, demonstrating usage of the model, and also making changes to the image generation for the output. Additionally, they contributed to the `operators` directory by adding code related to depth calculations using CUDA.
pytorchdeep-learningcnnscomputer-visionvideo-conversion
apache/mxnet
Oct 2015 - Jul 2018
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 & ML Engineer
Back-end Developer & ML EngineerContributions:10 releases, 589 commits, 2706 PRs in 2 years 8 months
Contributions summary:Eric's contributions primarily revolve around the development of the `mxnet/mxnet` repository, a flexible deep learning framework. Their work centers on enhancing the optimizer components, factory pattern utilization, multi-dimensional label support for the record I/O system, and also adding support for a new autograd function. The user demonstrated skills in coding with C++ to implement new operators like for the LSTM, including their integration with CUDA.
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.