When do I need to add or edit speaker names?
Marvin automatically detects the speakers on a call. In case you notice that some sentences in your call transcript are attributed to the wrong speaker or not attributed, you can fix this easily.
Navigate to the recording where you want to make the changes.
In the annotation tool, you'll have access to three options.
Assign speaker
Swap speaker
Unassign speaker
Assign speaker
Click on any speaker. You’ll be able to assign any of the call participants as this speaker. Select the correct participant name to proceed.
You’ll see a confirmation message and an option to update for just one instance or across the transcript.
You can also assign a new speaker who has not been detected. Click "Add new speaker".
Type the speaker name and choose whether you want to update this across the file.
Swap speaker
This option lets you interchange dialogues between any two call participants. For example, if you want to swap speaker A and speaker B, all speaker A dialogues will become speaker B dialogues and vice versa.
If you want to swap with a speaker who isn’t assigned yet, you must first attribute a part of the recording to them.
You’ll see a confirmation message to explain the change. Click "Swap speakers" to go ahead.
Unassign speaker
You can also unassign a speaker. Use the confirmation box to specify if the change should be across the transcript or only for that instance.
Edit participants
You can also add participants to the call that weren’t automatically detected. Click the “Edit participants” in the toolbar.
You can add a participant from your database or create a new one by typing out their name.
