Qiyuan Gong
AI Software Arch at Shanghai Innovation Institute
Minhang District, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolQiyuan Gong is an AI software architect with 12 years of experience building and optimizing large-scale ML and big data systems, currently focusing on LLM serving and fine-tuning for Intel AIPC at Shanghai Innovation Institute. He holds a PhD in Data Privacy from Southeast University and has deep expertise in privacy-preserving ML, federated learning, and PII/GDPR-related anonymization work that informs his production-grade deployment strategies. At Intel he led efforts across BigDL, OpenVINO, Cluster Serving and SSM/HDL, and more recently contributed to IPEX-LLM documentation that helps accelerate local LLM inference and fine-tuning on Intel XPUs. Practically minded and hands-on, he implements low-bit quantization, speculative decoding, MoE CPU/GPU hybrid inference, and operator-level optimizations like FlashAttention. Beyond core engineering he publishes code and algorithmic solutions on GitHub (including a multilingual LeetCode repo), blending research rigor with developer-centric documentation and tutorials. Based in Shanghai, he combines academic depth with production experience to bridge privacy research and high-performance LLM engineering.
12 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy (PhD), Data Privacy, Computer Science, Doctor of Philosophy (PhD), Data Privacy, Computer Science at Southeast UniversityBachelor's degree, Computer Science, Bachelor's degree, Computer Science at Xidian University
English, Chinese
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (25)
algorithms10
programming-language10
python10
lintcode10
technical-writing10
data-structure10
java10
javas10
data-structures10
documentation10
read-the-docs10
typed-array9
multidimensional-arrays9
dynamic-array9
arrayobject9
Programming languages (14)
Github contributions (5)
qiyuangong/leetcode
Mar 2016 - Mar 2022
Python & JAVA Solutions for Leetcode
Role in this project:
Full-stack Developer
Full-stack DeveloperContributions:77 reviews, 233 commits, 57 PRs in 6 years
Contributions summary:Qiyuan contributed to the development of LeetCode solutions, with implementations in both Python and Java, covering a range of problems. These contributions include solutions for common coding problems, and include code for data structures and algorithms. The user's contributions span both backend and front-end concepts, providing solutions in different languages, demonstrating proficiency in problem-solving. The user shows an understanding of core data structure implementation.
javapythonleetcode-solutionsleetcode
intel/ipex-llm
Apr 2019 - Nov 2022
Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, DeepSeek, Mixtral, Gemma, Phi, MiniCPM, Qwen-VL, MiniCPM-V, etc.) on Intel XPU (e.g., local PC with iGPU and NPU, discrete GPU such as Arc, Flex and Max); seamlessly integrate with llama.cpp, Ollama, HuggingFace, LangChain, LlamaIndex, vLLM, DeepSpeed, Axolotl, etc.
Role in this project:Technical Writer
Technical WriterContributions:652 reviews, 166 commits, 350 PRs in 3 years 8 months
Contributions summary:Qiyuan primarily contributed to the project by refining and expanding the documentation, specifically within the PPML (Privacy Preserving Machine Learning) section. Their work included linking documentation to the readthedocs page, fixing typos, reorganizing tutorials and menus, and adding new content such as the deployment guide and an attestation guide. These changes indicate a focus on improving the clarity, accessibility, and completeness of the project's documentation.
llm-inferencellama2pythonfinetuningllama
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.