AI compliance matching

Know what a document proves before you file it.

Quick Check lists the standards a policy or proof touches. Full Manual Analysis scores it against every requirement in your manual and quotes the supporting lines. On a standard, the AI Compliance Check shows what’s still missing.

  • Which standards a document touches
  • Every requirement scored, with quotes
  • Excel export and add to a standard
12 standards touched by one policy

Sound familiar?

Matching documents to standards is the slowest job in accreditation.

One policy can support a dozen standards. Finding every one, and the exact sentence each needs, is hours of reading.

  • “I read a 30-page policy to find the one sentence a standard needs.”

    Supporting lines are quoted and highlighted for you.

  • “I’m never sure I’ve found every standard a document supports.”

    Quick Check lists the standards it touches, by match level.

  • “I find out a subpoint isn’t covered when the assessor tells me.”

    Each requirement is rated, so gaps show up while there’s time.

Try it

One policy, every standard it touches.

This is a sample Quick Check for a body-worn camera policy. Filter by match level or sort by standard number, the way you would before deciding where the document belongs.

  • High 70% or above
  • Medium 40 to 69%
  • Low under 40%

Sample data, scripted on this page. Nothing is uploaded and no AI runs here.

Policy 412 Body-Worn Cameras.docx Sample Quick Check · 10 standards touched Sample

Showing 10 of 10 standards.

  1. 41.3.8 Body-worn camera use Subpoints matched: 4 of 4 95% High
  2. 42.2.1 Evidence handling Subpoints matched: 3 of 4 80% High
  3. 41.3.9 Recording retention Subpoints matched: 2 of 3 75% High
  4. 83.2.4 Digital evidence storage Subpoints matched: 2 of 4 55% Medium
  5. 17.5.3 Equipment issue and care Subpoints matched: 1 of 2 45% Medium
  6. 82.1.1 Records privacy Subpoints matched: 2 of 5 40% Medium
  7. 33.5.1 Training on new equipment Subpoints matched: 1 of 3 35% Low
  8. 26.1.1 Code of conduct Subpoints matched: 1 of 4 30% Low
  9. 1.2.9 Bias-free policing Subpoints matched: 1 of 5 20% Low
  10. 12.1.2 Chain of command Subpoints matched: 0 of 2 15% Low

Match levels: High 70% or above, Medium 40 to 69%, Low under 40%. Results are advice, not a compliance decision.

Feature · AI compliance matching

What it does, screen by screen.

Sample screens built from the real product: the same buttons, statuses and steps you’ll use, with sample data.

Full Manual Analysis

Every requirement in your manual, scored and quoted.

Full Manual Analysis checks a document against your whole manual. Each requirement gets a score and the quotes that support it, with a “Units Met” count, and you can add the document to a standard straight from the results.

  • Download the results as an Excel report
  • Run from the Document Library by an Accreditation Manager or Review Manager

AI Compliance Check

On a standard: what’s met, what’s partly met, what’s missing.

The AI Compliance Check rates every requirement of a standard Met, Partly Met, Not Shown or Missing, and gives an estimated compliance percentage. For Not Applicable and Optional requests it rates your justification Supported, Partly Supported or Not Supported.

  • Met counts as 1 and Partly Met as ½
  • Advice to act on, “not a compliance decision”

Use AI to Highlight

The right lines, tagged to the right subpoint.

On a standard, “Use AI to Highlight” marks the sentences in a policy or proof that meet each subpoint. Adjust them, drag to add your own and “Add Label”. The Covers panel shows what’s still uncovered.

Advice you can check

Built to be checked, not trusted blindly.

Full Manual Analysis shows the quotes behind every score, and an advisory notice says plainly that results are advice. AI runs on Amazon Bedrock with US-based processing.

  • 50 runs per agency per day by default
  • People set every status: your Review Manager, then the assessor
How we protect your data

Who uses it

The right people, with the right access.

These are the role names you’ll see in Badge 6. Each role sees and does only what it needs, and the rules are enforced on the server.

  • Agency

    Accreditation Manager

    Runs Quick Check and Full Manual Analysis, and adds documents to standards.

  • Agency

    Review Manager

    Runs the same checks and uses the ratings during internal review.

  • Agency

    Agency Admin

    Sees the AI Match % per standard on the Standards Progress Report.

  • Accrediting body

    Assessor

    Sees the highlighted lines on shared standards.

FAQ

Questions about AI compliance matching

Something else? Ask us directly.

Does the AI decide whether we’re compliant?

No. Every result is marked as advice, and the AI Compliance Check says it is “not a compliance decision”. Your Review Manager and, later, your assessor set the statuses.

What’s the difference between Quick Check and Full Manual Analysis?

Quick Check tells you which standards a document touches. Full Manual Analysis goes requirement by requirement through your whole manual, with scores, supporting quotes, a Units Met count and an Excel download.

How many checks can we run?

By default, 50 runs per agency per day.

Where does the AI run?

On Amazon Bedrock, inside AWS, with US-based processing. The security page has the details.

Can I change the highlights the AI made?

Yes. Remove them, drag to add your own, label them and tag them to subpoints. The highlights are yours.

This Is Accreditation: Simplified. This Is Badge 6.

Bring a policy. We’ll run it live.

In a 30-minute demo we’ll run Quick Check and a Full Manual Analysis on a document you choose, against your program’s standards.