John Kirkham
Senior System Software Engineer at NVIDIA
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolJohn Kirkham is a Senior System Software Engineer based in Seattle with 12 years of experience building high-performance, scientific and distributed systems—currently driving system software at NVIDIA. He is a prolific open-source contributor across the Python scientific stack and GPU ecosystem, with notable work in NumPy, Dask, CuPy, RAPIDS (rmm, cuML, cuDF) and core projects like Cython and OpenBLAS that improve numeric performance and interoperability. His strengths combine low-level performance engineering (CUDA/device buffers, BLAS optimizations) with robust backend design, testing and CI/DevOps automation for large projects such as conda-forge and Dask. He also contributes to data-format and tooling projects (zarr, h5py, zarr-python) that enable scalable scientific workflows and has added serialization and type-conversion features used in distributed contexts. A practical communicator and technical writer, he improves documentation and developer workflows—an unusual mix of performance tuning and developer experience refined by a physics background and a Stanford LEAD certificate.
12 years of coding experience
Certificate, LEAD Program, Certificate, LEAD Program at Stanford University Graduate School of BusinessMaster of Science (M.S.), Physics, Master of Science (M.S.), Physics at Emory UniversityB.S., Physics, B.S., Physics at Rhodes College
Stackoverflow
705reputation
53kreached
17answers
9questions
Github Skills (147)
cpython10
dependency-management10
python10
package-management10
utf10
testing10
data-manipulation10
bash10
ipython10
c1110
distributed-computing10
list-processing10
c1710
automation10
mat10
Programming languages (20)
Github contributions (5)
dask/dask
Feb 2015 - Nov 2024
Parallel computing with task scheduling
Role in this project:
Back-end Developer & Test Automation Engineer
Back-end Developer & Test Automation EngineerContributions:283 reviews, 399 PRs, 124 pushes in 9 years 10 months
Contributions summary:John primarily contributed to the `dask/dask` repository by implementing and improving core array operations. Their work included enhancing the internal workings of array functions, notably related to the shape and length, and adding testing for these functionalities. They also focused on improving NumPy compatibility within the library, which included support for edge cases in the `pad` operation, refactoring and testing core array components.
pythonschedulingparallelnumpydask
conda-forge/staged-recipes
Mar 2016 - Jan 2023
A place to submit conda recipes before they become fully fledged conda-forge feedstocks
Role in this project:DevOps Engineer
DevOps EngineerContributions:678 reviews, 2076 commits, 1910 PRs in 6 years 10 months
Contributions summary:John primarily contributed to the repository's infrastructure and build processes. Their commits focused on modifying the CI/CD configuration files, integrating with `conda-smithy` for feedstock management, and updating build scripts. They also addressed dependency management issues and implemented changes to improve build automation and consistency, specifically in the context of conda-forge recipes.
placeconda-forgerecipescondasubmit
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.