Sentinel RFP
Sentinel breaks an RFP into requirements and tasks, drafts responses, scores how well each claim is supported, and parks the result for a person to review.
How it works
Section titled “How it works”- Start a Sentinel run from RFP text.
- The workflow runs parse, plan, write and review steps, then scores trust.
- A trust scorer rates every drafted claim from 0 to 100. The rule is fixed and uses no AI: source authority (30 percent), claim support by text overlap (35), citation completeness (20) and source recency (15). A score of 80 or more is high, 50 or more is medium, and below 50 is low.
- Three review gates are checked: every extracted requirement is addressed by a drafted section, requirements that carry a FAR or DFARS reference are addressed, and every claim meets the run’s minimum trust score (50 by default).
- The run stops at awaiting review. A person records the review decision. Nothing is treated as done before that.
What runs today
Section titled “What runs today”AI drafting is off in production. With it off, the parse, plan, write and review steps return fixed sample responses instead of generated text. The trust scorer and the review gates are fixed rules and run the same way either way. Treat a run today as a demonstration of the workflow, not as drafted proposal text.
Why it matters
Section titled “Why it matters”Sentinel is built around the idea that drafts are inputs to review, not finished work. The gates are the point: automation drafts, humans approve.
Related
Section titled “Related”Try this in the platform