Voice profiles

Register known voices, with consent, so call and meeting transcripts show their name instead of Person A.

Where to find it

Voice › Voice profiles

Open this screen
Voice profiles page: setup state, the profiles table with their buttons, What is stored and Recognition settings

Steps

  1. In the sidebar, open Voice › Voice profiles.
  2. Check the setup card at the top: Ready or Needs setup for speaker labels and for voice recognition.
  3. Press Add a voice profile, fill in the name and type, tick the consent box and press Save with consent.
  4. Press Record a sample on the profile, record about 20 to 30 seconds or upload a file, and press Save voiceprint.
  5. Open a call or meeting and press Label speakers; known voices are shown by name.

Settings on this screen

SettingWhat it doesDefault
Add a voice profileA person's name, type (Client or Staff), an optional contact, and their consent.—
Record a sample / Add a sampleMakes a voiceprint from 20 to 30 seconds of the person speaking.—
Delete voiceprintDeletes the voiceprint; the name and consent record stay.—
Consent withdrawnRecords the withdrawal and deletes the voiceprint and the profile.—
Recognition settingsMatch threshold (1-100) and how many days transcripts with speaker labels are kept.70, 30 days
Search transcriptsFinds words or a name in transcripts with speaker labels.—
Consent recordsFor whom, recorded by, given, withdrawn, voiceprint deleted.—

Tips

  • The What is stored card lists exactly what is kept and for how long.

Good to know

  • Only workspace managers can add profiles or record samples.
  • Phone calls are labelled from their recording after the call; live labels need the operator's live speaker separation.

Related how-tos

Start free Home