William Chen
Co-Founder, Head Of Strategy & Management
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolWilliam Chen is a Beijing-based founder and engineering leader with nine years of cross-disciplinary experience spanning mechanical design, robotics, and machine learning systems. As Co-Founder and Head of Strategy at Sapient Intelligence he’s building a new AI architecture toward AGI, and previously led venture development for 20+ medical engineering projects at Tsinghua. His hands-on background ranges from LiDAR mechanical design at Hesai and RoboMaster robotics at DJI to co-founding a VTOL aircraft project, giving him rare end-to-end product instincts from hardware to AI. An active contributor to prominent open-source ML tooling—improving trainer performance and adding LoRA/Adapter integrations in well-known projects like OpenPrompt and OpenDelta—he blends systems optimization with practical model-tuning expertise. Colleagues describe him as relentlessly curious: he starts each day asking “What does the world need?” and turns that question into interdisciplinary prototypes and strategic initiatives.
9 years of coding experience
5 years of employment as a software developer
High School Diploma, High School Diploma at Cranbrook SchoolsBachelor's degree, Mechanical Engineering, Bachelor's degree, Mechanical Engineering at Tsinghua University
English, chinese-mandarin
Github Skills (15)
params10
code-optimization10
pytorch10
nlp10
para10
deep-learning10
optmization10
python10
optimisation10
natural-language-processing10
optimization10
deeplearning-ai9
transformer9
regular-expression8
lora8
Programming languages (9)
Github contributions (5)
thunlp/OpenPrompt
Sep 2021 - Dec 2021
An Open-Source Framework for Prompt-Learning.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:2 releases, 53 commits, 10 PRs in 2 months
Contributions summary:William primarily contributed to improving the OpenPrompt framework, specifically focusing on trainer and data processing functionalities. They fixed issues related to performance in the trainer, and configuration issues. They also added a beta tag to the project documentation, suggesting work on the project's overall build status and project release process. These changes involve the core framework code, indicating a focus on optimizing existing features and making new additions to the system.
natural-language-understandingpre-trained-language-modelsdialogue-systemsnatural-language-processingprompts
thunlp/OpenDelta
Feb 2022 - Nov 2022
A plug-and-play library for parameter-efficient-tuning (Delta Tuning)
Role in this project:ML Engineer
ML EngineerContributions:18 commits, 3 PRs, 14 pushes in 9 months
Contributions summary:William primarily contributed to the implementation and modification of parameter-efficient tuning techniques for pre-trained language models within the OpenDelta library. Their work involved fixing regular expression matching, refining documentation, and integrating LoRA, BitFit, and Adapter models. Additionally, the user made changes to the interactive modification features and updated structure mapping and bmtrain tutorials, demonstrating a focus on usability and integration with other libraries.
nlpdeltadeep-learningmachine-learningparameter
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.