Grant drafting support with guided generative AI software

Abstract

Grant writing is a time-consuming and labor-intensive process that is the essential mechanism by which most academic researchers fund their scientific investigations. To support this effort, we present a software application with corresponding source code that provides draft sections of a typical NIH-style grant to expedite the time to a first draft. The software provides draft grant text, affording researchers extra time to optimize the content and improve the probability of an award. There are three main components to the software: 1) a search and compare tool that compares the proposed research to previously funded NIH applications, 2) a specific aims page drafter, and 3) a research strategy section that incorporates significance, innovation, approach, rigor, and preliminary data. Our initial internal software deployment has been well received, and researchers indicate significant perceived time savings when using this tool.

Keywords

• Grant writing
• Generative ai
• Research strategy
• Specific Aims