Shenggui Li
Singapore
Summary
🤩
Rockstar🎓
Top SchoolShenggui Li is a computer scientist with seven years of hands-on experience building machine learning and full-stack systems across finance and tech, from Goldman Sachs reliability work to OCR and computer-vision solutions at YITUTech. He contributes to large-scale deep learning infrastructure—helping improve Colossal-AI’s AMP, tensor-parallel training, and 2D operation optimizations—demonstrating practical expertise in distributed training and convergence fixes. Comfortable across React, Java Spring, Kafka and deep learning frameworks, Shenggui blends research experience (Singtel CAI Lab, A*STAR) with production engineering, including a monitoring dashboard at Visa and sentiment analysis at Panasonic. Based in Singapore with a top-graded Computer Science degree from Nanyang Technological University, he pairs a pragmatic engineering mindset with curiosity—embodied by his “Stay Hungry, Stay Foolish” ethos—to make large AI models cheaper, faster and more accessible.
7 years of coding experience
3 years of employment as a software developer
Bachelor's degree, Computer Science, 4.88/5, Bachelor's degree, Computer Science, 4.88/5 at Nanyang Technological University
Chinese, English, French
Github Skills (10)
data-parallel10
pytorch10
deep-q-learning10
data-parallelism10
distributed-training10
deep-learning10
optimization10
ampl10
parallelization9
ai9
Programming languages (10)
Github contributions (5)
hpcaitech/ColossalAI
Nov 2021 - Jan 2023
Making large AI models cheaper, faster and more accessible
Role in this project:
ML Engineer
ML EngineerContributions:3 releases, 1561 reviews, 299 commits in 1 year 2 months
Contributions summary:Shenggui primarily contributed to the development and improvement of the Colossal-AI framework, a system for large-scale machine learning model training. Their work focused on enhancing the Torch AMP support within the framework for improved tensor parallel training, incorporating gradient accumulation, and fixing bugs. Furthermore, they optimized 2D operations, fixed convergence issues in various models, demonstrating expertise in distributed training and deep learning optimization.
heterogeneous-trainingcolossal-aifinetuningdeep-learninginference
FrankLeeeee/ColossalAI
Oct 2022 - Jan 2023
Colossal-AI: A Unified Deep Learning System for Large-Scale Parallel Training
Contributions:9 commits, 19 PRs, 1262 pushes in 3 months
pytorchcolossal-aiparalleldeep-learningmachine-learning
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.