The University of North Florida (UNF) has received a $105,000 grant from the Sloan Foundation to fund AI-assisted updates to legacy research software code. The project, based in UNF's School of Computing, aims to reduce the time researchers spend maintaining scientific software while preserving accuracy and reliability.
The Sloan Foundation, named after former General Motors President Alfred P. Sloan, supports scientific research through its Technology Program. The grant will fund work led by Associate Professor Upulee Kanewala and Director and Professor Nan Niu, with Graduate Research Assistants Eric Good and Nabin Chaulagain.
What the project targets
Researchers at UNF will use AI to support changes in the internal structure of legacy code as it gets updated. The goal is to make the programming easier to maintain and keep pace with evolving calculations. Many research teams around the world rely on software that has accumulated years of patches and modifications, making updates slow and error-prone.
The project also aims to produce evidence that AI testing can reduce technical barriers to maintaining research software. If successful, the approach could support long-term, large-scale software development focused on scientific research.
Why the Sloan Foundation funded the work
Joshua M. Greenberg, Director of the Sloan Foundation's Technology Program, said the grant addresses a core need in scientific computing.
"The durability and trustworthiness of scientific research depends on the rigor and robustness with which the underlying research code is built and maintained," Greenberg said. "This grant to UNF will help ensure that we can take advantage of new AI tools to safely restructure legacy research software as scientific needs evolve."
The work sits at the intersection of Generative Code and research infrastructure. Using AI to refactor legacy code is a growing area of interest for institutions that maintain software over decades rather than months.
Why this matters for science and research professionals
If you manage or contribute to research software, the UNF project is a test case for whether AI can handle structural code changes without introducing errors. The findings could inform how your own team approaches maintenance of legacy systems. The grant also signals that funders like the Sloan Foundation are willing to pay for AI-assisted software upkeep - not just new development. For researchers stretched across multiple projects, AI for Science & Research tools that reduce maintenance time could free up hours for actual scientific work.
Your membership also unlocks:
