Meetbook HelpWhat is voiceprint recognition?
Opt-in cross-meeting speaker ID explainer — what it does, that it's off by default.
Normally, speaker labels ("Speaker A", "Speaker B") only apply within a single meeting — the same person shows up as a fresh, unlabeled speaker every time they're in a different recording, until someone renames them. Voiceprint recognition changes that: once a voice has been identified once, Meetbook can recognize that same person automatically in future meetings, across your whole workspace.
It's opt-in and off by default
Because a voiceprint is biometric data, this feature is disabled by default for every workspace and has to be turned on deliberately by an admin — see Turning voiceprint recognition on or off. It's not something that activates automatically just because your workspace records meetings.
How it works, at a high level
The first time a speaker is identified by name in a recording (see A speaker is mislabeled — how to fix it), a voice signature is stored for them. In later meetings, new speakers are compared against stored signatures — a match gets auto-labeled with the known name instead of a generic "Speaker A".
What it's for
Mainly useful for workspaces with recurring external participants — the same clients, candidates, or partners showing up across many meetings — where manually renaming the same person every single time gets repetitive.
See Voiceprint & biometric data policy for how this data is handled and retained.