Andrew James
Director Of Engineering - Performance at OpenTeams
Chicago, Illinois, United States
Summary
🤩
Rockstar🎓
Top SchoolAndrew James is a Director of Engineering specializing in performance with 11 years of experience building high-performance backend systems and sparse tensor functionality for ML frameworks. Based in Chicago, he blends deep computational-science roots—pursuing PhD-level theoretical chemistry work—with production engineering, having contributed core features to PyTorch and quantum chemistry projects like Psi4. He has progressed from graduate research to senior and staff engineering roles and now leads performance engineering at OpenTeams, after senior technical leadership at Quansight. His open-source contributions include enabling advanced sparse formats, batched sparse operations, and excited-state solvers, showing a rare mix of numerical algorithm expertise and systems-level optimization. Colleagues rely on him to translate complex scientific algorithms into robust, GPU-accelerated implementations that scale. He’s equally comfortable refactoring C++/Python scientific code and driving cross-functional teams to ship measurable performance gains.
11 years of coding experience
11 years of employment as a software developer
Doctor of Philosophy - PhD (Incomplete) Theoretical Chemistry, Doctor of Philosophy - PhD (Incomplete) Theoretical Chemistry at Virginia TechBachelor of Science - BS Chemistry, Bachelor of Science - BS Chemistry at Marist University
Github Skills (23)
ec10
struct10
pytorch10
quantum-chemistry10
c-language10
python10
operation10
tensorrt10
sparse-matrix10
computational-chemistry10
numpy10
tensorflow10
sparse-data10
tensor10
cprogramming-language10
Programming languages (9)
Github contributions (5)
pytorch/pytorch
May 2022 - Jan 2023
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
Back-end Developer
Back-end DeveloperContributions:406 reviews, 409 commits, 81 PRs in 8 months
Contributions summary:Andrew contributed to the PyTorch library by implementing and improving sparse tensor functionality. Their work included enabling `index_add` for `ComplexHalf` data type, adding support for BSR <-> Strided Conversions, and incorporating the `spdiags` sparse matrix initialization. They also enhanced support for batch dimensions within sparse compressed formats, addressing a bug in BSR->Dense batched operations, and enabled CSC support for dense conversions. Furthermore, the user added support for `relu`, `log1p`, and `frac` in various sparse layouts and contributed to the `mm`/`addmm` operations.
pythongpu-accelerationdeep-learninggpunumpy
psi4/psi4
Jul 2015 - Jan 2019
Open-Source Quantum Chemistry – an electronic structure package in C++ driven by Python
Role in this project:Back-end Developer
Back-end DeveloperContributions:214 commits, 71 PRs, 58 pushes in 3 years 6 months
Contributions summary:Andrew's commits primarily involve modifications to the source code of the Psi4 quantum chemistry software package, specifically focusing on areas related to computational chemistry and electronic structure calculations. Contributions include the removal of example code from documentation, and the implementation of changes to natural orbital calculations within the dipole calculation modules. They appear to contribute with a focus on improving and maintaining the core features of the software package.
physicschemistrypythonquantum-computingelectronic-structure
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.