Skip to main content
Northdocs

Use cases

Redact the chat and the recording

For disclosure teams that receive Teams, Slack, WhatsApp, captions, or meeting audio and must ship a clean copy.

Litigation and municipal teams use Conversation Redact to produce chat exports and meeting recordings without leaving names, emails, CPR numbers, or speaker identity in the file. Review speaker turns, accept an identity once, then beep, silence, or forensically anonymize accepted audio. Chats count toward pages. Audio uses per-user voice minutes.

Start 30-day free trial See interactive demo

30-day free trial · no card required

Where conversation production fails

  • Chat exports mix people, IDs, and attachments across hundreds of turns.
  • A transcript wipe still leaves the same words in the recording.
  • A beep in an audio editor does not prove the name is gone. You ship the file and hope.

What Conversation Redact changes

  • Speaker-turn review with the same identity Fast Track as documents
  • Beep, silence, or forensic voice anonymize on accepted spans
  • Apply refuses the download if an accepted string remains in the transcript, or if a beep, silence, or anonymized span did not change the audio

How teams redact conversations

  1. 1

    Upload the export or recording

    Drop a Teams, Slack, or WhatsApp file, a VTT or SRT transcript, or audio. A zip may pair audio with captions. Audio without captions uses speech-to-text only when AI / ML detection is enabled; otherwise include a caption sidecar.

  2. 2

    Accept identities and choose audio treatment

    Review speaker turns. Accept or skip each identity. For recordings, pick beep, silence, or a forensic anonymizer strength.

  3. 3

    Apply, then download the clean copy

    Apply redaction or tokens. Download the rewritten transcript and, when present, the redacted WAV.

Questions for this use case

Is video supported?
No. Conversation Redact covers chats, captions, and audio. Upload a transcript or audio file, not a video container.
Which plans include it?
Trial, Starter, Pro, Firm, and Enterprise. It is on with the plan. Voice minutes are 60 per user on Trial and Starter, 180 on Pro, 600 on Firm, and unlimited on Enterprise. Outbound Guard seats do not include it.
Does speech-to-text require AI?
Yes. Speech-to-text runs only when AI / ML detection is enabled for the organization. If AI is off, upload a WebVTT or SubRip sidecar with the recording.

Start with a real pack

Upload documents, accept identities once, and export only after verification passes.

Start 30-day free trialGet the redaction checklist