Home

Reliable Agentic Workflows for Sustainable Multi-Language Scientific Software Interfaces

A URSSI Fellowship project (2026) investigating whether agentic AI workflows can reliably maintain and sustain cross-language interfaces for scientific software — using the Cantera MATLAB interface as a case study.

Central question: Under what conditions can agentic tools support release-quality maintenance of scientific software interfaces, and what recurring failure modes require automated validation, human review, or both?

• 📅 Project timeline
• 👤 About the author
• 💻 Source & artifacts on GitHub

---

Project blog

• Aug 1, 2026

  Can AI agents keep scientific software interfaces alive?

  The problem hiding in plain sight

subscribe via RSS