Shubham Kapoor
Senior Machine Learning Scientist at Amazon Web Services (AWS)
Berlin, Germany
Summary
🤩
Rockstar🎓
Top SchoolShubham Kapoor is a Senior Machine Learning Scientist based in Berlin with nine years of experience building production-ready ML systems at AWS and research institutions. He specializes in probabilistic time-series forecasting and has contributed notable improvements to the widely used GluonTS project, enhancing PyTorch estimators with LR schedulers, weight decay, custom imputation, and more robust deserialization. Across roles from applied research at Fraunhofer and RWTH Aachen to multiple machine learning scientist positions at AWS, he bridges research rigor with deployable engineering. Known for pragmatic model engineering, he focuses on turning advanced probabilistic models into reliable, maintainable services at scale.
9 years of coding experience
6 years of employment as a software developer
Master of Science (M.Sc.), Media Informatics, Master of Science (M.Sc.), Media Informatics at RWTH Aachen UniversityBachelor of Technology (BTech), Information Technology, Bachelor of Technology (BTech), Information Technology at Guru Nanak Dev Engineering College, Ludhiana
English, Hindi, Punjabi
Github Skills (11)
neural-network10
forecasting10
pytorch10
machine-learning10
forecast10
time-series-forecasting10
deep-learning10
python10
aws9
mxnet8
data-science8
Programming languages (3)
Github contributions (5)
awslabs/gluonts
Aug 2021 - Jan 2023
Probabilistic time series modeling in Python
Role in this project:
ML Engineer
ML EngineerContributions:29 reviews, 8 commits, 20 PRs in 1 year 5 months
Contributions summary:Shubham primarily contributed to the development and improvement of time series forecasting models within the GluonTS framework, specifically focusing on the PyTorch implementation. Their work includes adding features like learning rate and weight decay to estimators, integrating LR schedulers, enabling custom imputation methods, and adding functionality for deriving auto fields. Additionally, the user addressed code issues related to deserialization within the PyTorch predictor and updated mean scaler implementations.
forecastingpythontime-series-analysistimeseries-forecastingaws
shubhamkapoor/gluon-ts
Nov 2019 - Nov 2019
GluonTS - Probabilistic Time Series Modeling in Python
Contributions:2 PRs, 1 push in 1 day
forecastingdtwpythontime-series-analysismachine-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.