What is research-grant-finder-agent?

Research Grant Finder Agent is a Python tool that helps researchers find NIH funding opportunities worth applying to, instead of manually searching databases and reading long announcements one by one. Rather than doing a simple keyword search, it organizes the process into a workflow of several specialized AI agents, each handling one part of the job, from collecting opportunities to analyzing and ranking them by relevance. The idea behind the project is that a researcher searching for terms like data science may miss funding announcements that use different wording, such as predictive modeling, omics, or medical imaging, even though those opportunities are closely related to their work. The tool tries to catch those near misses by evaluating opportunities on several signals rather than one exact phrase match. Separate agents handle different sources and tasks: one pulls announcements from the NIH Guide, another searches Grants.gov, and a third gathers related project context from NIH RePORTER. A parser agent converts all of these different formats into one common structure, duplicates are filtered out, and a relevance agent scores each opportunity using computational keyword signals, whether relevant terms appear in the title itself, similarity to known AI and data-science grants, whether the source is NIH or HHS, upcoming deadlines, and how complete the listing's information is. Opportunities are then grouped into high, medium, and watchlist priority tiers. Results are stored in a SQLite database and can be reviewed through a Streamlit dashboard, exported to Excel, or summarized in a digest report. There is also a feedback agent meant to capture researcher input so relevance scoring can be tuned over time. The project describes itself as a research prototype built with Python 3.10 or later.