Why we built this feature
It may be possible that our speech recognition system did not match correctly the participant and the speech (audio and transcript).
Hence, we built the functionality to manually edit the voice samples of the participant to reprocess the recording and produce a correct match.
How it works
1. Visit the meeting page.
2. In the participants component, click the edit "pencil" icon.
3. For each participant, you can edit the voice sample. Make sure the voice sample time frame includes the audio of the participant.
4. Click "Reprocess"
5. After reprocessing, the participants' diarization and insights will be improved as a result of using the voice sample references.
