What it is
Nonprofits spend a great deal of time finding grants they might win. Grant Match AI scores funders against a nonprofit's profile so the search starts with the best matches.
The decision
Zen Aegis split the score. Four of the six factors (geography, budget, eligibility and prior relationship) are computed by a deterministic, tested scorer. Only mission and population fit go to a model, batched into one structured call to keep cost down.
The code is public.
What was built
- Built a six-factor match score: mission, geography, budget, eligibility, population served and prior relationship.
- Built a three-step onboarding, a dashboard, sign-in by emailed link and chat for each organization.
- Seeded funders from public nonprofit filings and a curated grant list.
- Wrote unit tests for the scoring engine.
Results
- Phase 1 is public on GitHub, created April 10, 2026.
- Phase 2 is planned: a pipeline board, drafted grant narratives, a story bank, a document vault and billing.
Public record
At a glance
| Project type | Web application |
|---|---|
| Industry | Nonprofit |
| Seat | Founder, Zen Aegis |
| Ran through | Zen Aegis |
| Dates | April to June 2026 |
| Related pages | Zen Aegis |
| Credits | Common Ground Robby Prochnow, Founder, Zen Aegis: designed and built it |