Danni Li
Machine Learning Engineer at Rime Labs
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolDanni Li is a Machine Learning Engineer based in the San Francisco Bay Area with eight years of experience building ML systems and contributing to production-grade software at companies from Meta to Tesla and Amazon. She currently works at Rime Labs after an AI Resident role at Meta, and has a strong internship track record across major tech and research labs including Microsoft Research Asia and ByteDance. Danni is an active open-source contributor to foundational ML projects like PyTorch and ONNX, where she implemented vmap support, GRU gate improvements, and shape/quantization checks—work that influences both developer ergonomics and model interoperability. Her background blends research and production: she has research internships and full-stack roles, plus a BS in Computer Science from City University of Hong Kong and study stints at several global universities. Colleagues describe her as someone who moves fluidly between low-level library internals and applied model engineering, often spotting edge-case failures before they reach production.
8 years of coding experience
2 years of employment as a software developer
University College LondonNanyang Technological University SingaporePeking UniversityBachelor of Science - BS Computer Science, Bachelor of Science - BS Computer Science at City University of Hong Kong
Japanese, Chinese, English
Github Skills (15)
quantization10
pytorch10
machine-learning10
inference10
onnx10
deep-learning10
shapes10
python10
autograd9
tensorflow9
neural-network9
c-language8
cprogramming-language8
documentation7
numpy7
Programming languages (13)
Github contributions (5)
onnx/onnx
May 2023 - Jan 2025
Open standard for machine learning interoperability
Role in this project:
ML Engineer
ML EngineerContributions:12 reviews, 8 PRs, 20 comments in 1 year 8 months
Contributions summary:Danni primarily contributes to the ONNX repository by addressing issues related to model checking, shape inference, and quantization. They've implemented exception checks within the `check_model` function and exposed `LexicalScopeContext` in `checker.py`. Additionally, the user modified the code to print tensor dtypes as strings in shape inference, and added a default context in the check_function. Their work also includes documentation updates and addressing quantization related issues.
pytorchmxnetdeep-learninginteroperabilitymachine-learning
pytorch/pytorch
Dec 2022 - Dec 2022
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:Back-end Developer
Back-end DeveloperContributions:15 reviews, 1 commit, 39 PRs in 1 day
Contributions summary:Danni primarily contributed to the PyTorch library, focusing on improvements and additions to core functionalities. Their work involved implementing vmap support for various operations like `tril`, `triu`, and `linalg.lu_factor`, adding support for the new gate calculation in GRU, and updating documentation. They also addressed issues by modifying existing code and test files and made updates to include `nn.ParameterDict` in dynamo `__getitem__`
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.