Nathan Cooper
Research Scientist at Answer.AI
United States
Summary
🤩
Rockstar🎓
Top SchoolNathan Cooper is a research scientist with a decade of experience applying software engineering and machine learning to real-world problems, currently advancing retrieval-augmented models at Answer.AI. He holds a PhD in Computer Software Engineering from William & Mary and a strong academic track record dating back to a 3.84 GPA in his software engineering undergraduate work. An active open-source contributor, Nathan has improved DeepMind-inspired RETRO implementations—fixing reconstruction, device handling, and generation issues—demonstrating attention to robustness and reproducibility in retrieval-enhanced transformers. Colleagues describe him simply as “a nerd,” which captures his deep curiosity and hands-on approach to debugging complex ML training pipelines. He pairs rigorous research training with practical engineering discipline to move prototype models toward reliable production behavior.
10 years of coding experience
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at College of William and MaryBachelor's degree, Software Engineering, 3.84, Bachelor's degree, Software Engineering, 3.84 at University of West FloridaAssociate of Arts (A.A.), General Studies, Associate of Arts (A.A.), General Studies at Pensacola State CollegeDoctor of Philosophy - PhD, Computer Software Engineering, Doctor of Philosophy - PhD, Computer Software Engineering at William & Mary
English, Spanish
Stackoverflow
331reputation
9kreached
9answers
4questions
Github Skills (15)
transformers10
pytorch10
machine-learning10
deep-learning10
python10
debug9
attention-mechanism9
debugging9
json6
datetime6
bash6
indexing6
android6
adb6
aws-amplify6
Programming languages (15)
Github contributions (5)
lucidrains/RETRO-pytorch
Apr 2022 - May 2022
Implementation of RETRO, Deepmind's Retrieval based Attention net, in Pytorch
Role in this project:
ML Engineer
ML EngineerContributions:2 reviews, 7 commits, 4 PRs in 5 days
Contributions summary:Nathan primarily contributed to bug fixes and improvements within the RETRO-pytorch repository, which focuses on retrieval-enhanced transformers. Their commits demonstrate a focus on refining the core functionalities of the retrieval and training modules, specifically addressing issues related to reconstruction errors, device handling, and generation processes. They also made adjustments to the dependency management and code style, by fixing a comma issue and updating the version, as well as saving stats to avoid recomputation.
pytorchtransformersretroattention-mechanismdeep-learning
CarperAI/squeakily
Nov 2022 - Jan 2023
A library for squeakily cleaning and filtering language datasets.
Contributions:1 review, 23 commits, 9 PRs in 2 months
nlpcleaningdatasetmachine-learningdatasets
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.