Meet Shah
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMeet Shah is a Research Engineer with 12 years of experience building and deploying production-scale ML systems for autonomous systems and multimodal perception, currently at Google DeepMind after roles at Waymo, Uber ATG, and Facebook AI Research. He specializes in speeding and compressing models and has a strong track record integrating pre-trained vision-and-language checkpoints and improving data pipelines, including notable contributions to the popular mmf framework (adding BAN, GQA support, and reproducible checkpoint metadata). Trained at IIT Bombay (B.Tech + M.Tech EECS), he blends deep research instincts with practical engineering to move models from experiments into latency- and resource-constrained production. Based in Mountain View, he favors pragmatic, reproducible solutions that scale under tight timelines and often focuses on the overlooked engineering work—data prep, tooling, and metadata—that makes research deployable.
12 years of coding experience
2 years of employment as a software developer
Indian Institute of Technology Bombay
English
Github Skills (10)
pytorch10
deep-learning10
multimodal10
python10
pre-trained-model10
model-checking9
checkpointing9
checkpoint9
computer-vision9
nlp8
Programming languages (7)
Github contributions (5)
facebookresearch/mmf
Sep 2018 - Jun 2019
A modular framework for vision & language multimodal research from Facebook AI Research (FAIR)
Role in this project:
ML Engineer
ML EngineerContributions:18 commits, 2 PRs, 5 pushes in 9 months
Contributions summary:Meet made several contributions to the `mmf` repository, focusing on the integration and use of pre-trained models, particularly in the context of Vision and Language tasks. Their work included updating feature extraction scripts, modifying download paths for datasets, uploading and updating pre-trained model paths, and adding support for the GQA dataset, indicating a focus on data preparation and model utilization. The user also implemented a new model (BAN) and added features for reproducibility, such as adding git metadata and the current experiment config to checkpoints.
pytorchmodular-frameworkdialogvisionvision-language
jck/riscv
May 2016 - Aug 2016
Contributions:100 commits, 6 PRs, 8 comments in 2 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.