Docs › See who said what (speaker labels)

See who said what (speaker labels)

Pandecho tells participants apart automatically and separates you from the room, so quotes and action items are attributed correctly.

Speaker diarization runs on-device as part of transcription. It labels each line by speaker and distinguishes your mic from the meeting audio — which is what lets Pandecho tag action items to the right owner.

In the app

1. Record any meeting — diarization is automatic, nothing to enable.
2. In the transcript, each line shows a speaker label (e.g. Speaker 1 (You)).
3. Rename or merge speakers in the transcript if you want cleaner labels.

Ask your AI agent

Agents receive the speaker labels in the transcript. Set this up once with the MCP guide, then try:

• Summarize what each participant committed to in my last meeting.
• List everything the customer (not me) asked for.