Jing Xu
Manager at Fujitsu
Tokyo, Japan
Summary
🤩
Rockstar🎓
Top SchoolJing Xu is a technical leader and manager with eight years of hands-on experience optimizing PyTorch and building AI solutions, now based in Tokyo. She has driven Intel’s PyTorch support, co-designed the Intel Extension for PyTorch, and integrated VTune/ITT profiling into the upstream PyTorch codebase—work that influenced widely used performance tooling and tutorials on pytorch.org. Fluent in Japanese, English, and Chinese, she combines customer-facing consulting across the full lifecycle with deep engineering skills in LLMs, RAG, Hugging Face Transformers, and large-scale profiling. Known for translating customer feedback into product fixes, automation tools, and clear documentation, she also regularly delivers live trainings and workshops at major events. Pragmatic and quick to adopt new tech, she pairs PhD-level AI training with practical open-source contributions that improve performance on Intel platforms.
7 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy - PhD Coursework, Artificial Intelligence, Doctor of Philosophy - PhD Coursework, Artificial Intelligence at Tokyo Institute of TechnologyBachelor's degree, Electrical and Electronics Engineering, Bachelor's degree, Electrical and Electronics Engineering at Nanjing University of Science and Technology
English, Chinese, Chinese, Japanese
Github Skills (24)
pytorch10
performance-analytics10
performance-monitor10
c-language10
intel10
python10
machine-learning10
performance-measurement10
performance-analysis10
deep-learning10
optimisation10
performance-tuning10
performance-monitoring10
cprogramming-language10
optimization10
Programming languages (7)
Github contributions (5)
intel/intel-extension-for-pytorch
Apr 2021 - Jan 2023
A Python package for extending the official PyTorch that can easily obtain performance on Intel platform
Role in this project:
ML Engineer
ML EngineerContributions:19 releases, 19 reviews, 232 commits in 1 year 9 months
Contributions summary:Jing primarily focused on optimizing the installation process and dependencies for the Intel extension for PyTorch, a package designed for performance on Intel platforms. They modified the `setup.py` file to handle dependencies, manage torch wheel file packaging, and address potential build issues. Further contributions included fixing runtime errors, cleaning the installation folder structure, and addressing issues related to the compilation of the C++ SDK.
pytorchpythondeep-learningintelmachine-learning
pytorch/pytorch
Aug 2021 - Oct 2022
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:ML Engineer
ML EngineerContributions:99 reviews, 14 commits, 37 PRs in 1 year 2 months
Contributions summary:Jing's contributions focused on integrating Intel's VTune Profiler's Instrumentation and Tracing Technology (ITT) APIs into PyTorch for performance profiling. This involved adding ITT API calls to annotate PyTorch operations and custom code scopes, similar to NVIDIA's NVTX. Furthermore, the user implemented and tested on-demand verbosing functionality for oneMKL and oneDNN, allowing for detailed operator execution information and execution time profiling. The user also updated documentation and added unit tests related to ITT functionality.
pythongpu-accelerationdeep-learninggpunumpy
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.