Piotr Bialecki
Director Of Engineering at NVIDIA
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolPiotr Bialecki is a Director of Engineering at NVIDIA with 11 years of experience building and scaling ML and deep learning infrastructure from research to production. Based in the San Francisco Bay Area, he combines people leadership with hands-on contributions to high-profile open source projects like PyTorch and NVIDIA/apex, optimizing mixed-precision training and multi-CUDA build pipelines. He progressed through technical and managerial roles at NVIDIA after a background as a freelance ML developer and academic researcher, bringing both product-facing and research sensibilities. Known for pragmatic automation and CI/CD work, he has repeatedly modernized build and test workflows to support the latest CUDA toolchains across platforms. An MS in Information Technology and a BS in Biomedical Engineering underpin his interdisciplinary approach to ML systems engineering.
11 years of coding experience
11 years of employment as a software developer
Bachelor of Science - BS Biomedical/Medical Engineering, Bachelor of Science - BS Biomedical/Medical Engineering at Mannheim University of Applied Sciences, Heidelberg UniversityMaster of Science - MS Information Technology, Master of Science - MS Information Technology at Mannheim University of Applied Sciences
Github Skills (26)
pytorch10
github-ci10
docker10
python10
machine-learning10
conda10
bash10
dockers10
cicd10
multiprecision10
script10
deep-learning10
gpu10
ci-cd-pipeline10
build-automation10
Programming languages (7)
Github contributions (5)
pytorch/builder
Aug 2020 - Jan 2023
Continuous builder and binary build scripts for pytorch
Role in this project:
DevOps Engineer
DevOps EngineerContributions:13 reviews, 24 commits, 43 PRs in 2 years 5 months
Contributions summary:Piotr primarily contributed to the project's build and deployment infrastructure, specifically focusing on CUDA versions and their integration within the pytorch/builder repository. Their work involved modifying build scripts (install_cuda.sh, conda/build_all_docker.sh) and configuration files to support different CUDA versions (11.0, 11.6, 11.7, 11.8). They also updated the `conda/pytorch-nightly/build.sh` for the nightly builds, and various other changes related to dependencies and build configurations.
pytorchbuild-scriptscontinuousbuilder
NVIDIA/apex
Dec 2018 - Apr 2021
A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch
Role in this project:ML Engineer
ML EngineerContributions:14 reviews, 24 commits, 111 PRs in 2 years 4 months
Contributions summary:Piotr primarily focused on updating and optimizing example code within the `nvidia/apex` repository, which provides tools for mixed precision and distributed training in PyTorch. Their contributions involved adapting example scripts for newer PyTorch versions (>=0.4.0), and making targeted changes, such as correcting code, removing unused tensors, and fixing CUDA related issues. They also contributed to the addition of a DCGAN example, and made improvements for the use of AMP.
pytorchraymixed-precisiondeep-learningtemporal-data
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.