Yicong Du - Software Engineer at Meta

Yicong Du

Software Engineer at Meta

Mountain View, California, United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Yicong Du is a software engineer with a decade of experience building scalable AI infrastructure and performance-focused systems, currently at Meta in Mountain View. He previously drove performance and architecture work at Airtable and contributed to AI metadata, TorchEval, and feature store initiatives at Facebook. His open-source contributions include meaningful refactors to the widely used PyTorch Lightning codebase—streamlining core infrastructure, migrating to native fsspec APIs, and adding type hints for long-term maintainability. Comfortable across back-end systems and AI tooling, he focuses on pragmatic cleanup and durable platform improvements that reduce technical debt. Trained at Carnegie Mellon with an MS in Electrical and Computer Engineering, he blends rigorous engineering fundamentals with production-grade system design. Colleagues describe him as a steady contributor who surfaces practical improvements that keep complex ML systems running smoothly.

10 years of coding experience

4 years of employment as a software developer

Master of Science - MS Electrical and Computer Engineering, Master of Science - MS Electrical and Computer Engineering at Carnegie Mellon University

Github Skills (9)

pytorch10

machine-learning10

fsspec10

deep-learning10

pytorch-lightning10

python10

ai10

data-science10

refactoring9

Programming languages (1)

Python

Github contributions (5)

Lightning-AI/pytorch-lightning

Feb 2022 - Mar 2022

Pretrain, finetune ANY AI model of ANY size on multiple GPUs, TPUs with zero code changes.

Role in this project:

Back-end Developer

Contributions:60 reviews, 20 commits, 31 PRs in 1 month

Contributions summary:Yicong primarily contributed to refactoring and removing functionalities within the PyTorch Lightning framework. They updated the codebase to use the native fsspec API for file system interactions and added type hints to improve code readability. Furthermore, the user removed several deprecated properties and functionalities from the `AcceleratorConnector`, indicating a focus on streamlining and maintaining the core infrastructure of the project.

pythonheadachespytorch-modelsdata-sciencehandling

DuYicong515/pytorch-lightning

Jan 2022 - Mar 2022

The lightweight PyTorch wrapper for high-performance AI research. Scale your models, not the boilerplate.

Contributions:169 pushes, 44 branches in 2 months

pytorchpythonboilerplatedeep-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.

Request Free Trial