Proposal Review demo

Marks each claim in a draft by what your records say, traces each requirement and drafts the fixes 1:34

What the demo shows

Checks a draft proposal for unsupported claims, missed requirements and inconsistencies

The draft proposal is shown as the customer will read it, with every factual claim tinted: green where a source supports it, amber where nothing does, red where a source contradicts it, with the count for each paragraph in the margin. An intake agent reads the draft, the RFP and the compliance matrix; a claim agent finds every figure, commitment, certification and reference; an evidence agent tests each one against the SLA schedules, contract playbook, certificate and project registers, price book and past proposals to the same buyer; a requirements agent traces every requirement to its paragraph; a consistency agent compares numbers and names across sections and past bids; a checker agent re-checks every red without seeing the first working; a fix agent drafts each fix from the source wording. The proposal lead talks to the agents beside the draft and approves, edits or rejects each fix with a reason; legal wording goes to the legal reviewer; delivery accepts the service commitments and the VP signs the gold-team gate.

Transcript

  1. Proposal Review checks a draft proposal against your own records before the customer sees it.
  2. 79% of proposal teams now draft with AI, checking it is still manual.
  3. Havergate's $14.6 million bid, 318 claims, checked overnight.
  4. Green is supported, amber has no source, red is contradicted.
  5. The draft promises $99.99 percent uptime.
  6. The service schedule says 99.9.
  7. A checker agent agrees.
  8. The fix comes from the source.
  9. Proposal lead Noah Fisher approves it with a reason.
  10. The agents also caught another customer's name in Section 3.
  11. All 146 requirements are traced.
  12. 7 need work. Before gold team, 27 commitments with owners, then the vice president signs.
  13. Every claim cited, every decision signed, every agent step logged.
  14. The target, zero contradicted promises sent, and red team in one day, not three.
  15. Built on Botminds: agents, data, screens, and governance on one platform.

Transcribed from the film’s narration. Captions are on the player.