Check strategic fit against your stated priorities, with quotes

Fit factors come from your own priorities

List the factors you fund on: issue area, geography, population served, program type. Cemented AI checks each proposal against that list the same way every time. No house definition of "fit" is imposed on your portfolio.

Every factor is quoted from the application

For each fit factor, the relevant passage is pulled from the proposal and quoted. You see where the applicant addresses your geography or population, in their own words. Quotes are checked against the source document.

The alignment call goes to a person, with the evidence attached

Whether a proposal aligns with your mission is a judgment, so the model never decides it. The factor and the quoted application text are routed to a program officer. The reviewer reads the evidence and makes the call. Every subjective decision works this way.

Gaps are flagged, not guessed

If a proposal never addresses a fit factor you named, that gap is flagged as unaddressed. The agent does not infer a population or invent a geography to fill the space. A missing answer reads as missing.

Declined applicants get a reason traceable to a stated priority

When a proposal is declined on fit, the reason ties back to a priority you published. It also ties to the passage it was weighed against. The decision is defensible to your board and explainable to the applicant.

How a review runs

1. Step 1

   Name your fit factors

   Enter the priorities you fund on: issue area, geography, population, program type. These become the checklist every proposal is read against.

2. Step 2

   Run the deterministic checks

   Each proposal is checked against every factor the same way. Where it addresses a factor, the passage is quoted and verified against the source.

3. Step 3

   Route the alignment call to a reviewer

   The factor and the quoted text go to a program officer. The reviewer reads the evidence and decides whether the proposal aligns. The model does not decide it.

4. Step 4

   Record a traceable decision

   Each fit finding traces to the passage it came from and the priority it was weighed against. Declined applicants get a reason tied to a stated priority.

Common questions

Does the model score alignment?

No. The model never scores subjective fit. Deterministic checks confirm whether the proposal addresses each factor you name and quote the passage. The alignment judgment is routed to a program officer holding that evidence.

What if your priorities are nuanced and hard to reduce to a checklist?

The deterministic checks only confirm whether each factor is addressed and surface the quote. Nuance is exactly what goes to your reviewer. The point is to hand a person the relevant passage, not to replace the judgment with a number.

How does this reduce bias in fit calls?

Each proposal is checked against the same published factors, in the same order, with the supporting passage quoted. Reviewers decide alignment from the application text rather than from impression. The basis for each call is recorded and consistent.

What happens when a proposal does not mention one of your priorities?

That factor is flagged as unaddressed. The agent will not guess a population, geography, or program type the applicant did not state. The gap is shown plainly for your reviewer to weigh.

Can this connect to your grants management system?

Connectors to systems like Submittable, Fluxx, or Foundant are handled on Enterprise. Ask about connectors and the setup that fits your intake.

How do you get started and what does it cost?

See the pricing page for plans, or reach out to talk through your fit factors and intake process. Bring your published priorities and a set of past proposals to see how the checks and quotes read against real applications.