Kollol Das
Machine Learning Engineer at Sensibill
Old Toronto, Ontario, Canada
Summary
👤
SeniorKollol Das is a Machine Learning Engineer with nine years of experience specializing in document extraction using state-of-the-art generative AI. Based in Old Toronto, he combines independent research instincts with practical engineering, shipping robust sequence models and production-ready solutions. He has contributed to the widely used tensor2tensor library, implementing LSTM-based seq2seq models with Bahdanau and Luong attention and resolving stability issues like shape mismatches and NaN losses. Comfortable across research and engineering, Kollol focuses on making advanced models reliable and maintainable in real-world pipelines. His background suggests a knack for debugging tricky numerical problems and translating cutting-edge techniques into dependable systems.
9 years of coding experience
Github Skills (12)
attention-mechanism10
sequence-to-sequence10
machine-learning10
lstm10
rnn-model10
deep-learning10
tensorflow10
n10
python10
machine-translation9
reinforcement-learning3
tpu3
Programming languages (2)
Github contributions (5)
tensorflow/tensor2tensor
Jun 2017 - Oct 2017
Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research.
Role in this project:
ML Engineer
ML EngineerContributions:13 commits, 4 PRs, 16 comments in 3 months
Contributions summary:Kollol primarily contributed to the development of LSTM-based models within the tensor2tensor library, focusing on sequence-to-sequence tasks. Their work included implementing LSTM with attention mechanisms, including both Bahdanau and Luong attention variants, and integrating these into the existing seq2seq framework. Furthermore, they addressed and resolved shape mismatch issues and NaN loss problems, demonstrating a focus on the model's functionality and stability.
pytorchautoencoderdeep-learningmachine-translationreinforcement-learning
kolloldas/Namehacker
Nov 2017 - Sep 2018
Contributions:14 commits, 6 pushes, 1 branch in 9 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.