Adrish Dey
Doctoral Student at Boston University
Boston, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolAdrish Dey is a PhD student in Computer Science at Boston University with eight years of industry and research experience at the intersection of geometry, topology, and machine learning. He has blended applied engineering roles—at Weights & Biases and Rephrase.ai—with open-source contributions to major projects like geomstats (PyTorch backend improvements and autograd support) and TensorFlow Datasets (Mozilla Common Voice integration). His research work spans geometric problems in graphics and ML, including contributions during MIT summers on quad-mesh optimization and geometric statistics. Adrish brings practical MLOps and backend skills along with a mathematical bent, able to translate advanced manifold computations into robust, tested code. He welcomes collaborations with concise research proposals and often focuses on numerical robustness and datatype correctness in geometric ML implementations.
8 years of coding experience
1 year of employment as a software developer
Bachelor of Technology - BTech, Computer Science and Engineering, Bachelor of Technology - BTech, Computer Science and Engineering at Netaji Subhash Engineering CollegeDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Boston University
English, Bengali, Hindi
Stackoverflow
1reputation
636reached
1answer
1question
Github Skills (19)
pytorch10
geometry10
back-end-development10
python10
machine-learning10
audio-processing10
geo10
tensorflow10
tensorflow-datasets10
geom10
data-engineering9
data-processing9
autograd9
csv9
deep-learning8
Programming languages (10)
Github contributions (5)
tensorflow/datasets
Mar 2019 - Mar 2020
TFDS is a collection of datasets ready to use with TensorFlow, Jax, ...
Role in this project:
Back-end Developer & MLOps Engineer
Back-end Developer & MLOps EngineerContributions:70 commits, 13 PRs, 177 comments in 1 year
Contributions summary:Adrish primarily worked on implementing the Mozilla Common Voice dataset within the TensorFlow Datasets library. They added the necessary files and code to load and process the dataset, including defining the dataset's features and splits. Their contributions involved integrating the dataset with the TensorFlow ecosystem. They also made improvements to the code to handle data types and correct errors.
datanumpydeep-learningdatasetmachine-learning
geomstats/geomstats
Oct 2021 - Dec 2021
Computations and statistics on manifolds with geometric structures.
Role in this project:Back-end Developer
Back-end DeveloperContributions:12 commits, 5 PRs, 5 comments in 1 month
Contributions summary:Adrish primarily contributed to the backend of the `geomstats/geomstats` repository, focusing on the PyTorch backend. Their work involved refactoring and improving the `logm` function, including adding an adjoint method for autograd and expanding data type support. They also addressed data type issues within the Frechet mean implementation, and updated eigvalsh with the official PyTorch implementation and added unit tests for logm.
gpu-programmingstatisticslie-groupsdeep-learningneural-networks
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.