Grant Finder
A small scheduled grant-discovery service across six regions, with deterministic scoring, deduplication, and emails only for new matches.
What it helps with
A solo operator can lose hours checking several national and regional sources, then receive the same opportunities again the next day. The useful automation is a dependable shortlist, not an invented claim that the organisation is eligible.
What was built
The service checks official APIs where they exist, public pages where needed, and search only where a better source is unavailable. It normalises the results, removes duplicates with stable identifiers and titles, applies a visible profile score, stores what it has already seen, and sends email only for new matches above the threshold. One failed source does not hide the state of the others.
What made it different
The project is deliberately lean. It avoids scraping sources that offer an API, avoids repeated or empty emails, and keeps the scoring simple enough to inspect. The operator still reads the grant and verifies eligibility before spending time on an application.
A grant score can help order the reading list, but it is not an eligibility decision.
What it demonstrates
- Six-region scheduled discovery
- Stable deduplication and deterministic profile scoring
- No repeat or empty emails; visible source failures