Prabhat Roy
Machine Learning Engineer at Meta
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolPrabhat Roy is a Machine Learning Engineer with nine years of experience building end-to-end ML systems across NLP, Computer Vision and classical ML, currently at Meta in London. He has a strong track record at Microsoft and Facebook driving model conversion and deployment (notably contributions to ONNX Runtime and sklearn-onnx) and leading TorchVision video initiatives that added GPU video decoding and new datasets. As a founder of eLiterate.me and PlayCraft Games, he blends product thinking with teaching and hands-on engineering, having built an e-learning platform and shipped an image-based quiz game on Google Play. His open-source work spans low-level C/C++ backend changes and high-level model converters, demonstrating fluency from runtime internals to ML pipeline interoperability. Comfortable leading small teams and coordinating large cross-functional efforts, he also brings a subtle focus on robustness—fixing segfault-causing tensor issues and improving converter test coverage.
10 years of coding experience
12 years of employment as a software developer
Master's degree Computer Science, Master's degree Computer Science at University of Southern CaliforniaIntermediate Mathematics and Computer Science, Intermediate Mathematics and Computer Science at Chinmaya Vidyalaya, Bokaro Steel CityBachelor's degree Computer Science, Bachelor's degree Computer Science at The National Institute of Engineering, Mysore
English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (27)
pytorch10
adaboost10
c-language10
python10
feature-selection10
testing10
scikit10
machine-learning10
audio-processing10
capi10
onnx10
datasets10
keras10
onnxruntime10
lightgbm10
Programming languages (7)
Github contributions (5)
onnx/sklearn-onnx
Dec 2018 - Jul 2019
Convert scikit-learn models and pipelines to ONNX
Role in this project:
ML Engineer
ML EngineerContributions:31 commits, 221 PRs, 132 pushes in 6 months
Contributions summary:Prabhat primarily contributed to the development of converters for scikit-learn models to ONNX format. Their work includes implementing converters for various feature selection transformers, polynomial features, K-bins discretizers, AdaBoost classifiers, and nearest neighbor algorithms. They also addressed issues in existing converters, such as those for calibrated classifiers and scalers, improving overall functionality and compatibility within the project.
onnxscikit-learn
pytorch/vision
Apr 2021 - Apr 2022
Datasets, Transforms and Models specific to Computer Vision
Role in this project:ML Engineer
ML EngineerContributions:223 reviews, 247 commits, 169 PRs in 11 months
Contributions summary:Prabhat significantly contributed to the implementation of the KITTI dataset, including the creation of the Kitti dataset class, data parsing logic, and documentation. They also ported the SVHN dataset to a new test framework and added tests for STL10 and Places365 datasets. Additionally, the user addressed floor_divide deprecation warnings and added download support and test for the KITTI dataset, demonstrating a focus on expanding and improving dataset functionality.
computer-visionmachine-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.