Reviewing a Draft Proposal

Recently updated

Use AI to review your grant proposal for compliance, quality, and completeness before submission.

Last updated Mar 24, 2026

The old way

You print out the RFP requirements, go through your draft with a highlighter, try to check off each criterion manually. You miss something — a page limit, a required attachment, a formatting requirement buried on page 12. Your colleague reads it for tone but doesn’t know the RFP well enough to catch compliance gaps. You submit and hope for the best.

The Grantable way

Use the /review skill and the AI runs a structured three-pass review — checking compliance against every RFP requirement, evaluating narrative quality, and flagging anything that’s missing or incomplete. It cites specific issues with specific evidence so you know exactly what to fix.

Step by step

1. Get your draft ready

Make sure your draft proposal is in your file tree. If you wrote it in Grantable using /write, it’s already there. If you wrote it externally, upload the document (Word or PDF).

Also make sure the RFP or application guidelines are uploaded — the AI needs both your draft and the requirements to run a thorough review.

2. Run the /review skill

Start a chat and invoke the review:

  • /review “Review our draft proposal against the RFP requirements”
  • /review “Check this LOI for completeness and tone”
  • /review “Does our budget narrative match the budget table?”

3. Review the findings

The AI runs three passes:

Compliance check — Does the draft address every requirement in the RFP? Are page limits met? Are all required sections present? Are attachments and forms accounted for?

Narrative quality — Is the writing clear and compelling? Are claims supported by evidence? Is the tone appropriate for this funder? Are there gaps in the logic or argument?

Completeness — Is anything missing? Are there sections that feel thin or underdeveloped? Are all budget items justified?

For each finding, the AI cites the specific issue, where it is in the draft, and what needs attention.

4. Fix the issues

Work through the flagged items. You can ask the AI to help with revisions directly:

  • “Strengthen the evaluation plan section — the reviewer said it’s thin”
  • “Add specific outcome data to support the claim in paragraph 3”
  • “Rewrite the methodology to address the timeline requirement”

5. Run a final check

After making revisions, run /review one more time to confirm everything is addressed. This final pass catches anything you might have introduced during editing.

Tips

  • Review early, not just at the end. Run /review on individual sections as you draft them, not just on the complete proposal.
  • Share the review results with your team. The findings make great discussion points in a team review meeting.
  • Don’t ignore low-confidence findings. When the AI flags something but isn’t certain, it’s often worth checking — those edge cases are where missed requirements hide.
  • Pair with the dashboard. Use the dashboard calendar view to make sure you’re running reviews with enough time before the deadline.

Related

© 2026 Grantable. All rights reserved.