Founder OS · Internal Proof of Concept

Matching every founder to the capital they qualify for — automatically

We already collect everything needed to connect founders to non-dilutive grants. It just sits static on a form. This turns that data into a ranked, per-founder shortlist the program team can work.

Built: July 2026  ·  Tested on: 2025 Spring cohort (frozen data)  ·  For: Dakota & the program team  ·  No live systems touched

The proof

90
real founders run through the matcher
62
grants in the internal database, matched on eligibility
44
founders (of 90) with an actionable, best-fit grant
2.7
avg best-fit grants per matched founder — a real shortlist, not a data dump

How it works

01 / INTAKE

Data we already have

Onboarding fields — demographics, stage, business age, revenue, location — the founder already gave us.

→
02 / MATCH

Eligibility, checked

Each founder is tested against every grant's real rules. Identity, revenue floors and ceilings, business age, geography.

→
03 / SHORTLIST

Ranked by fit

A tiered list per founder. Anything we can't verify is flagged for a human — never asserted.

What the team actually sees

Sample FounderIllustrative example · Nashville, TN
WomanFounder of colorVeteranDisability
Best fitGoogle for Startups Black Founders Fundup to $150,000
Best fitNAACP Powershift Entrepreneur Grant$25,000
Best fitBlack Girl Ventures$5K–$10K
Best fitBusiness Impact Grant (BIG)$5,000
Best fitHerRise MicroGrant$1,000/mo
Plus flagged for review: veteran and disability status also open veteran- and multi-group grants — surfaced separately so staff confirm the niche eligibility before acting. The tool never guesses.

The four tiers

Best fitMeets the grant's specific target + real dollars. Act on these.
EligibleSingle-identity match with a deadline or amount.
Standing resourceAlways-open programs (SBA, MBDA). Same for everyone.
Needs reviewConstraint we can't verify — sent to a human, not asserted.

Where it lives & what's next

Lives in ClickUp — where you already work

Not a new app to learn. Matches land in a Grant Matches list in the program workspace. The team updates status, shares grants, clears flags — independently, no dev in the loop.

The matcher only writes the machine fields. Your notes and status are never overwritten.

Next steps

  • Review these results together — do the matches hold up to your judgment?
  • Clean up three form fields before fall intake (adds grant-seeking + industry)
  • Point it at the fall cohort as acceptances land
  • Then: Project Healthcare federal-grant scraper