Kai Yao
City of Edinburgh, Scotland, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolKai Yao is a deep learning engineer and researcher with six years of experience spanning academia and industry, currently based in Edinburgh. He has driven model optimization and acceleration at Intel—building tooling like Neural Coder, contributing PyTorch adapters for Intel Neural Compressor, and benchmarking open-source AIGC models for CPU/GPU performance—while earlier research roles applied CNNs and microfluidics to single-cell morphology at Johns Hopkins. His background bridges cyber security PhD work, mechanical engineering, and applied ML for communications (5G MU‑MIMO), giving him unusual cross-domain fluency from low-level kernel tuning to biological imaging. An active contributor to the widely used intel/neural-compressor project, he focuses on quantization, mixed precision, and integration with Hugging Face tooling to make state-of-the-art compression practical.
4 years of coding experience
6 years of employment as a software developer
Doctor of Philosophy - PhD, Cyber Security, Privacy and Trust, Doctor of Philosophy - PhD, Cyber Security, Privacy and Trust at The University of EdinburghDiploma of Education, Architectural Engineering, Diploma of Education, Architectural Engineering at Osaka UniversityJohns Hopkins UniversityBachelor of Science - BS, Theoretical and Applied Mechanics, Bachelor of Science - BS, Theoretical and Applied Mechanics at Fudan University
English, Japanese, Chinese, German, Korean
Github Skills (11)
quantization10
pytorch10
machine-learning10
optimization10
large-language-models9
huggingface9
benchmark8
benchmarking8
auto-tuning8
p87
sparse7
Programming languages (4)
Github contributions (5)
intel/neural-compressor
Jul 2022 - Jan 2023
SOTA low-bit LLM quantization (INT8/FP8/INT4/FP4/NF4) & sparsity; leading model compression techniques on TensorFlow, PyTorch, and ONNX Runtime
Role in this project:
ML Engineer
ML EngineerContributions:5 reviews, 88 commits, 33 PRs in 6 months
Contributions summary:Kai primarily contributes to the integration and optimization of machine learning models within the `intel/neural-compressor` repository. Their commits involve adding and modifying code to support various quantization techniques, including INT8, BF16, and the integration with Hugging Face's Optimum-Intel framework. They also focus on enabling benchmarking and accuracy evaluation for different optimization strategies. The user's work also includes adding and improving the JupyterLab extension to support the neural compressor tool.
knowledge-distillationauto-tuningcompressorsparsityintel
kaikaiyao/kaikaiyao.github.io
Apr 2023 - Mar 2025
Contributions:122 pushes in 1 year 11 months
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.