Skip to content

DraftGuard Report Builder

AI police report writing.
Officer-led review.

Prepare an editable narrative from your notes and evidence. Review the facts, refine the report, and download PDF and Word files with the supporting report history.

Part of the DraftGuard investigative suite. Built by a detective.

From notes and evidence to a signed report

Supply the facts, review the narrative, and finalize the report.

  1. 01 / Supply

    Start with your material

    Type or paste notes into the Report Builder. Add supported BWC audio, dashcam recordings, photos, PDFs, or other documents; processed evidence text gives the draft additional context.

  2. 02 / Review

    Edit the draft

    Review the generated narrative, answer follow-up questions, and edit the report. Verify names, dates, quoted statements, and the distinction between observations and inferences.

  3. 03 / Export

    Finalize and download

    The certification workflow captures an officer signature. Download a ZIP with PDF and Word report files and a review manifest, then preserve the appropriate records in your agency’s systems.

Usable outputs

A draft you can inspect.
Files you can take with you.

The report narrative is editable before finalization. PDF and Word exports include the report and signature, AI interaction log, and evidence manifest. The ZIP also contains a review manifest with supporting review metadata.

BWC reports and the first AI draft

For reports drafted from body-worn camera evidence, DraftGuard preserves the first completed AI-generated draft while the officer edits and reviews the working narrative. PDF and Word report exports include the AI interaction history, including that first draft, alongside the report narrative.

A small review example

Fictional training notes. No real case data.

Example input / typed notes

At 14:10, I arrived at a fictional bicycle theft call. The reporting party said the bicycle was last seen at 12:30. I observed an empty bicycle rack. I did not witness the bicycle being taken. No suspect was identified in these notes.

What the officer should check

  • Attribute the last-seen time to the reporting party.
  • Distinguish the officer's observation from the reported theft.
  • Check that the narrative contains only facts supported by the source notes.
  • For a covered California report, confirm disclosure, program identification, and review records before finalization.

Put Report Builder into your workflow

What should officers review before finalizing?

Verify the narrative against the source material. Check names, dates, quotations, and attribution; correct errors and resolve missing details before completing the officer certification.

What should a California agency establish?

Use an agency policy that covers AI disclosure, program identification, officer review and signature, first-draft preservation, and audit records. The California SB 524 guide provides the requirements and a practical checklist.

How can an agency request access?

Submit an agency pilot request to discuss report inputs, review procedures, records needs, and current commercial terms.

See the workflow in your agency’s context

Start the discussion with your report inputs, officer review process, and records requirements.