Evan Smothers
Research Engineer at Netflix
Greater Seattle Area United States
Summary
🤩
Rockstar🎓
Top SchoolEvan Smothers is a research engineer at Netflix with nine years of industry experience building and deploying machine learning systems. He spent six years at Meta contributing to PyTorch and TorchMultimodal, where he improved multimodal model support, refactored for type safety, and modernized CI/CD and dependency management. His open-source work includes implementing a LoRA linear layer and establishing GitHub Actions pipelines for PyTorch tooling, bridging model development with robust DevOps practices. Earlier roles in data science at Uber and internships in industry and academia underpin a strong applied-math and production ML background. He holds a Ph.D. in Mathematics from UC Davis and a dual BS/BA in Mathematics and Music from the University of Rochester, signaling both rigorous quantitative skill and creative breadth. Notably, he tends to operate at the intersection of research-quality models and production-ready infrastructure, making experimental ideas ship-ready.
9 years of coding experience
13 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Mathematics, Doctor of Philosophy (Ph.D.), Mathematics at University of California, DavisBachelor of Science (B.S.), Bachelor of Arts (B.A.), Mathematics, Music, Bachelor of Science (B.S.), Bachelor of Arts (B.A.), Mathematics, Music at University of Rochester
English, Spanish
Github Skills (17)
transformers10
pytorch10
github-ci10
integrate10
python10
machine-learning10
integrations10
cicd10
transformer-models10
lora10
githubaction-workflow10
data-integration10
refactor9
typehinting9
test-automation9
Programming languages (3)
Github contributions (5)
pytorch/torchtune
Mar 2024 - Apr 2025
PyTorch native post-training library
Role in this project:
Back-end Developer & DevOps Engineer
Back-end Developer & DevOps EngineerContributions:2 releases, 1537 reviews, 337 PRs in 1 year
Contributions summary:Evan's primary contribution focused on setting up and configuring a continuous integration and continuous delivery (CI/CD) pipeline using GitHub Actions. They established the initial workflow by setting up the necessary infrastructure for the project. Additionally, the user implemented a LoRA (Low-Rank Adaptation) linear layer, and integrated a test script for validating its correctness. These changes are focused on model training and configuration aspects of the project.
facebookresearch/multimodal
Mar 2022 - Nov 2022
TorchMultimodal is a PyTorch library for training state-of-the-art multimodal multi-task models at scale.
Role in this project:Back-end Developer & MLOps Engineer
Back-end Developer & MLOps EngineerContributions:391 reviews, 78 commits, 93 PRs in 8 months
Contributions summary:Evan primarily contributed to the TorchMultimodal library, focusing on enhancing its capabilities for multimodal deep learning models. They removed support for older Python versions, updated dependencies, and refactored modules for type checking, emphasizing code quality and maintainability. The user also implemented support for MDETR models, including both the image and text encoders and the core model class, indicating a focus on integrating and optimizing cutting-edge multimodal architectures.
pytorchmulti-taskartdeep-learningmultimodal
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.