Methodology and framework for integrating AI agents into scientific software development workflows.
This project develops practical guidance for scientific software teams adopting AI coding agents. It addresses a gap in the community: while AI tools are rapidly being adopted, there is little principled methodology for integrating them responsibly into scientific workflows where correctness, reproducibility, and domain expertise matter.
Author: Victor Weeks, NSF National Center for Atmospheric Research (NSF NCAR)
- Framework — Principles and patterns for responsible AI-agent integration in scientific software
- Tutorials — Hands-on guides with interactive examples (coming October 2026)
- Community Catalog — Curated collection of tools, patterns, and case studies (coming October 2026)
- Workshops — Recordings and materials from community events (coming April 2027)
This work is supported by the Better Scientific Software Fellowship Program, funded by the U.S. Department of Energy and the National Science Foundation.