Keith Kraus
Director Of Engineering - CUDA Python And CUDA C Core Libraries at NVIDIA
New Jersey, United States
Summary
🤩
Rockstar🎓
Top SchoolKeith Kraus is a Director of Engineering at NVIDIA leading CUDA Python and CUDA C++ core libraries, with a decade of experience building high-performance GPU software and developer ecosystems. He co-founded Voltron Data and previously grew RAPIDS from zero into a sizable open-source GPU analytics community, acting as an architect for libraries like cuDF and RMM and launching CUDA-Python. Skilled across CUDA, C++, Python, Cython, Conda, and CI/CD, he combines hands-on build and automation work—evident from contributions to BlazingSQL and cuDF—with strategic product and engineering leadership. Based in New Jersey, he has scaled teams from small startups to 30+ engineers and aligned proprietary and open-source roadmaps to land customers. Notably, he’s a conda-forge core member and blends low-level systems expertise with developer experience focus to make GPU computing more accessible.
10 years of coding experience
13 years of employment as a software developer
Master of Engineering (MEng) Networked Information Systems, Master of Engineering (MEng) Networked Information Systems at Stevens Institute of Technology
Stackoverflow
241reputation
13kreached
10answers
0questions
Github Skills (26)
python10
cd10
bash10
conda10
cicd10
ci-cd-pipeline10
build-automation10
pipeline10
build-pipelines10
dockers9
docker9
rapids9
cuda8
documentation8
gpu-acceleration8
Programming languages (16)
Github contributions (5)
rapidsai/cudf
Mar 2018 - Aug 2022
cuDF - GPU DataFrame Library
Role in this project:
Back-end Developer
Back-end DeveloperContributions:1492 reviews, 496 commits, 1417 PRs in 4 years 5 months
Contributions summary:Keith primarily focused on code style improvements, including fixing PEP8 violations in tests and documentation. This suggests a focus on code quality and maintainability. The user also made changes to address an issue with a linebreak in the documentation, as well as adding and fixing copyright headers. The commits indicate a minor code maintenance role.
cudadataframe-librarydata-analysiscppcudf
BlazingDB/blazingsql
Sep 2020 - Oct 2020
BlazingSQL is a lightweight, GPU accelerated, SQL engine for Python. Built on RAPIDS cuDF.
Role in this project:DevOps Engineer & Build Automation Engineer
DevOps Engineer & Build Automation EngineerContributions:6 reviews, 14 commits, 3 PRs in 12 days
Contributions summary:Keith primarily focused on maintaining and updating the build process and CI/CD pipeline for the BlazingSQL project. They made changes to the `build.sh` script, conda recipes, and CI/CD scripts (`ci/cpu/upload_anaconda.sh`) to update dependencies, re-enable features, and ensure successful uploads to the appropriate conda channels. These contributions involved integrating and building various dependencies, and managing the deployment process.
cudfgpu-accelerationpythondata-sciencegpu
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.