Mannat Singh
AI Research Engineer at Meta
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolMannat Singh is an AI Research Engineer at Meta with a decade of experience applying machine learning and systems engineering to real-world problems across finance and industry. Trained at IIT Bombay and Caltech in electrical engineering with a CS minor, he transitioned from quantitative strategy at Goldman Sachs to AI roles at Bloomberg and now FAIR, combining strong mathematical foundations with production-focused research. At Meta he contributes to both model research and engineering, and as a DevOps contributor to the well-known facebookresearch/ClassyVision project he improved CI/CD, linters, and tooling to make large-scale vision experiments more reproducible. Colleagues rely on him for bridging research prototypes and robust deployment pipelines, and he has a knack for uncovering subtle infrastructure bugs that affect model training at scale.
10 years of coding experience
3 years of employment as a software developer
Indian Institute of Technology BombayCalifornia Institute of Technology
Github Skills (12)
sh10
github-ci10
script10
shell10
githubaction-workflow10
ci-cd-pipeline10
scripting10
testing5
dockers5
python5
git5
docker5
Programming languages (2)
Github contributions (5)
facebookresearch/ClassyVision
Sep 2019 - Apr 2022
An end-to-end PyTorch framework for image and video classification
Role in this project:
DevOps Engineer
DevOps EngineerContributions:2 releases, 6 reviews, 223 commits in 2 years 7 months
Contributions summary:Mannat primarily focused on improving the CI/CD pipeline and general project setup. They fixed an issue in the formatting script to correctly check formatting and implemented fixes for the linter's behavior within CircleCI, specifically addressing problems with the Git diff and environment variable settings. Furthermore, the user updated the isort settings, pinned the version, and integrated the model complexity hook to test a bug related to sync batch normalization.
pytorchimagenetresnetend-to-endmeta-learning
facebookresearch/maws
Oct 2023 - Mar 2025
Code and models for the paper "The effectiveness of MAE pre-pretraining for billion-scale pretraining" https://arxiv.org/abs/2303.13496
Contributions:1 review, 7 PRs, 18 pushes in 1 year 5 months
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.