Projects
Large-scale open-source tools I lead or contribute to. Each built to serve research communities, handling codebases of 100K+ lines across multi-institutional teams.
Lead Developer · 205+ GitHub stars · Docs
Python library for unstructured climate grid analysis. Climate scientists working with next-generation grids (MPAS, ICON, SAM) lacked tools for conservative analysis that preserve integral quantities across non-uniform meshes — a gap blocking petabyte-scale research.
I lead development since inception: core mathematical operators including conservative zonal averaging via Gauss-Legendre quadrature, Grid I/O for multiple formats (ESMF, MPAS, SCRIP, HEALPix), full CI pipeline, and regular PyPI releases. Currently building an MCP server and AI agent for natural-language interaction with climate datasets. Adopted by NCAR, DOE labs, and universities worldwide.
Aurora Exascale System · Argonne ALCF
PyTorch-based reimplementation of the Pangu-Weather deep learning framework for climate modeling, using the Spectral Fourier Neural Operator (SFNO). Ported and optimized for DOE's Aurora supercomputer — the first U.S. exascale system — with 60,000+ Intel GPUs. Demonstrates feasibility of AI-driven weather prediction beyond NVIDIA ecosystems, advancing DOE's mission for exascale Earth science.
Core Contributor · R&D 100 Award 2023
Hyperparameter optimization for cancer drug response prediction at supercomputer scale. Built the HPO infrastructure and ran 10,000+ training experiments across Summit, Theta, and Cori. Developed GitHub Actions workflows for cross-study validation. The 15+ researcher multi-lab collaboration (Argonne, LLNL, ORNL) relied on my benchmarking framework. Published in Briefings in Bioinformatics (2025). Additional papers: CANDLE/Supervisor, Counterfactuals.
I/O & Compression Lead · R&D 100 Award 2022 · Paper
I/O optimization for a million-line exascale multiphysics simulation engine used by hundreds of researchers for astrophysics, combustion, and fluid dynamics. Checkpoint/restart was consuming 30–50% of runtime on leadership-class supercomputers. I implemented asynchronous HDF5 I/O with Argobots and integrated SZ3/ZFP compression — achieving 40–70% reduction in checkpoint times on Summit and 50%+ storage savings. Enabled cross-checkpoint restart between AMReX and Paramesh solvers (a first for FLASH). Published at SC24 DRBSD-10.
MeshKit
Principal Investigator · DOE NEAMS (2009–2016) · Source
Open-source C++ toolkit for automated nuclear reactor core mesh generation. As PI, led design and development of lattice hierarchy-based meshing, parallel generation capabilities, and multi-format I/O. Adopted by reactor simulation teams at Argonne. Won Best Paper Award at the International Meshing Roundtable (2010).
Technical Expertise
Languages: Python, C++, Fortran, R, Bash, SQL
ML & Data: PyTorch, TensorFlow, NumPy, Pandas, Xarray, Scikit-learn, Parsl, Swift/T
HPC & Systems: MPI, OpenMP, HDF5, NetCDF, MOAB, Docker, Singularity, GitHub Actions
Domains: Climate modeling, cancer pharmacogenomics, computational physics, mesh generation, AI/ML infrastructure, reproducible workflows
Selected Publications
22+ publications · Full list on Google Scholar
Benchmarking community drug response prediction models
Partin, A., ..., Jain, R., et al. · Briefings in Bioinformatics, 27(1), 2025
Enabling Data Reduction for Flash-X Simulations
Jain, R., Tang, H., Dhruv, A., Byna, S. · DRBSD-10 Workshop, SC24
Cross-HPO: Optimizing Neural Networks for Cancer Drug Response
Jain, R., Wozniak, J.M., Partin, A., et al. · CAFCW24 Workshop, SC24
Creating Geometry and Mesh Models for Nuclear Reactor Core Geometries
Tautges, T.J., Jain, R. · Journal of Engineering with Computers, 2011
Recognition
EB-1A Extraordinary Ability
U.S. permanent residency granted under the EB-1A classification for extraordinary ability in sciences — reserved for individuals with sustained national or international acclaim.
CANDLE — Cancer Distributed Learning Environment for drug response prediction. The "Oscars of Innovation."
FLASH-X — Multiphysics simulation software for exascale computing.
ATPESC Scholar, 2015
Argonne Training Program on Extreme-Scale Computing — competitive program for HPC researchers.
Best Paper Award, 2010
International Meshing Roundtable — reactor core mesh generation.
Graduate Fellowship, 2007
Arizona State University — research assistantship in structural and computational mechanics.
Service & Mentorship
- SBIR/STTR Proposal Reviewer — U.S. Department of Energy
- Panelist — 5th Infraday Midwest Event ("Revolutionizing Public Infrastructure with AI")
- Reviewer — Journal of Open Research Software, NumGrid
- Committee — NumGrid 2020 Program Committee Member
Mentored several students and doctoral candidates over the years on research software engineering, HPC techniques, and open-source development practices.
Background
Argonne National Laboratory 2009 – present
Principal Specialist, Research Software Engineering. Lead developer for UXarray, FLASH-X, CANDLE/IMPROVE, MeshKit, and urban simulation projects. Division of Mathematics and Computer Science.
University of Chicago 2023 – present
Staff At-Large. Joint appointment supporting cancer pharmacogenomics and earth science research.
Arizona State University 2007 – 2009
Research & Teaching Assistant. Structural and Computational Mechanics Lab. Research on blast mitigation via FEM-based design optimization.
Education
M.S. Computer Science — University of Chicago (2020)
M.S. Structural Engineering — Arizona State University (2009)
B.Tech Mechanical Engineering — IIT ISM Dhanbad (2006)
Google Scholar · Resume · Full CV
