Arun Maiya
Computer Scientist at Institute For Defense Analyses, Inc.
Washington DC-Baltimore Area United States
Summary
🤩
Rockstar🎓
Top SchoolArun Maiya is a computer scientist at the Institute for Defense Analyses with seven years of experience applying machine learning, NLP, computer vision, and network science to extract meaning from raw data. He is the creator of open-source tooling such as ktrain, CausalNLP, and OnPrem.LLM, aiming to make deep learning more accessible and deployable. His hands-on contributions to ktrain include refining critical components like the learning-rate finder, reflecting deep practical knowledge of model training dynamics. Based in the Washington D.C. metro area, he bridges research-grade methods with pragmatic software engineering for defense and public-sector applications. Colleagues know him for turning complex computational problems into usable libraries and reproducible workflows. He mixes academic rigor with a pragmatic focus on usability, often prioritizing developer ergonomics in open-source ML tools.
7 years of coding experience
University of Illinois at Chicago
Github Skills (9)
algorithm10
machine-learning10
deep-learning10
python10
optimizers10
optimisation10
optimization10
tensorflow9
keras9
Programming languages (3)
Github contributions (5)
amaiya/ktrain
Apr 2019 - Jan 2023
ktrain is a Python library that makes deep learning and AI more accessible and easier to apply
Role in this project:
ML Engineer
ML EngineerContributions:166 releases, 2852 commits, 38 PRs in 3 years 10 months
Contributions summary:Arun made contributions focused on implementing, and refining the find_called method, which is pivotal to the functionality of learning rate finders for deep learning models. They also fixed an issue with the implementation of find_called and merged changes from the develop branch suggesting that they play an active role in the ktrain project, potentially contributing to or maintaining the learning rate finder related parts of the library. These contributions demonstrate a deep understanding of the learning rate optimization used for machine learning model training.
pythonktraintensorflowseq2seqcomputer-vision
amaiya/onprem
Aug 2023 - Apr 2025
A tool for running on-premises large language models with non-public data
Contributions:59 releases, 12 PRs, 558 pushes in 1 year 7 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.