Tag speakers in a conversation to distinguish who is speaking and train Otter to automatically identify speakers in future conversations.
Speaker Identification
Otter will automatically tag a speaker if it can, as well as group paragraphs of the same speakers. The paragraph grouping is automatic.
By tagging a speaker, you will teach Otter to recognize that speaker and improve upon the speaker identification for your future conversations, leading to less tagging in the future.
- Shared speakers
- Automatically tagged speakers
- New speakers
- Unknown speakers
- Manually tagged speakers
Shared speakers
For users part of a Workspace, the member’s speaker profile can be automatically used by Otter to tag them in a conversation when you have a meeting with them.
For users that have shared a conversation with users in a Channel, future conversations in that Channel can utilize the speaker tags within that Channel.
Example 3: Workspace members’ speaker identification profiles are shared.
A Workspace has the members Soraya, Armin, and Kaori. A 4th member, Luna, was added later to the Workspace. Each of their speaker tags for their own profile is shared across the Workspace.
Soraya starts a meeting with Armin. In Soraya’s conversation, Armin is automatically tagged, since he is in the same Workspace as Soraya, based on his speaker tag data.
Example 2: A Channel member records a conversation where a shared speaker is recognized
Bob belongs to the "Sales Team" Channel and adds "Jenny" as a speaker to a conversation that is shared with the “Sales Team” Channel. Another member of the “Sales Team” Channel records a conversation where “Jenny” is speaking. Otter will automatically recognize “Jenny” as a speaker and tag her name to the sections of the conversation where she spoke.
Example 3: A non-Channel member records a conversation with the same speaker
Alice does not belong to the “Sales Team” Channel and records a conversation where “Jenny” is speaking. Otter will not add her name to the sections of the conversations where she spoke. Alice will need to tag “Jenny”.
Automatically tagged speakers
Otter uses the existing speakers you have in your account and shared speakers in your Workspace to automatically identify speakers in a conversation.
New speakers
For conversations where a speaker has never been tagged before, Otter will tag them as Speaker # (i.e., Speaker 1, Speaker 2, Speaker 3, etc.). This indicates that Otter was able to distinguish a unique speaker throughout the conversation, but not associated with any current identified speaker in your Otter account. Tagging these speakers will automatically tag all other associated Speaker #s in the conversation.
Unknown speakers
If Otter was not able to identify a unique speaker, it may show as an unknown speaker . You can tag these speakers to start building the speaker identification for that speaker.
Manually tagged speakers
You can manually tag speakers to help Otter learn the speaker’s voices and identification. Each time you tag a speaker, Otter will be able to recognize that speaker in future conversations, so you have less tagging to do in the future.
A green check mark will appear next to speakers to indicate that the speaker was manually tagged.
Tag a speaker
You can tag a speaker to begin building the voice identification for that speaker.
- Navigate to the conversation.
- Click on the Transcript tab.
- Click on the speaker icon (e.g., Speaker 1). A pop-up will appear with a list of suggested speakers. You can tag your speaker in a variety of ways:
- Click on a suggested speaker
- Search for an existing speaker
- Create a new speaker.
- Otter will begin tagging and confirming the new speaker tag. All speakers with the same identified # will be tagged as that speaker. Wait until the speaker has completed.
- You will see a green checkmark next to the speaker once the speaker tagging process has been completed.
Otter will do its best to accurately tag speakers; however, the may be cases where you need to retag a speaker.
- Click on the speaker you want to retag.
- You can retag your speaker in a variety of ways:
- Click on a suggested speaker
- Search for an existing speaker
- Create a new speaker
- Click on the new speaker to confirm.
- Wait for Otter to finish confirming the new speaker.
You can rename a speaker anytime and it will update that speaker’s tag across all conversations.
- Click on any speaker icon to open the speaker list menu.
- Search and find the speaker you want to rename.
- Click on the 3-dot menu next to the speaker. Select Rename.
- Type the new speaker name and click the blue checkmark to confirm the changes.
If you have past conversations where you have not tagged the speakers yet and they still show as Speaker # (e.g., Speaker 1, Speaker 2, Speaker 3, etc.), but you know the speaker was tagged in a new conversation, you can rematch past conversations to tag those speakers with the new speaker data. Otter will use the recent most up-to-date speaker identification data in your account or Workspace to rematch those speakers.
- Navigate to a conversation where you have a Speaker #.
- Click on the 3-dot menu in the upper right corner.
- Select Rematch.
- Otter will use your account or Workspace’s present-day speaker data to tag any Speaker #s with a matched speaker.
Note: This is only for speakers that show as speaker #. Otter will not be able to rematch unknown speakers.
Example:
- Deborah records a meeting on February 1st, 2024. Deborah, Dani, and Gustavo spoke in the meeting. After the conversation finished processing, she noticed Otter was able to automatically tag the speaker Deborah, but not Dani and Gustavo, as it was the first time she had recorded a conversation with them.
- Otter tags them (what would be Dani and Gustavo) as Speaker 1 and Speaker 2 respectively. Deborah decides to not tag them with names at that time.
- Deborah later records a meeting on March 5th, 2024 with Dani and Gustavo again. This time she tags them as Dani and Gustavo.
- Now that Dani and Gustavo have a speaker identification profile, Deborah goes back to the February 1st conversation and clicks Rematch.
- Otter tags Speaker 1 and Speaker 2 as Dani and Gustavo respectively due to the new speaker identification data.
You can untag a speaker anytime to disassociate that audio from that speaker. We recommend retagging the speaker instead of untagging it so that Otter can associate the correct speaker with the audio.
- Click on the speaker you want to untag.
- Click Untag.
- Otter will untag that speaker.
Note: Otter will only untag that paragraph the speaker is associated with. If there are places in the transcript where the speaker was tagged incorrectly, you will need to untag them there as well. If you continue to see incorrect tagging issues, please contact our Otter Support Team.
FAQs
- No. You cannot delete a speaker at this time.
- Our team is working on adding speaker management in the future.
- In the meantime, if you have any speaker management issues, please contact our Otter Support Team.
- No. The Rematch feature will only tag the speaker identified as a Speaker # (e.g. Speaker 1, Speaker 2, Speaker 3, etc.)
- You will need to manually retag existing tagged speakers.
- No. The Rematch feature will only tag the speaker identified as a Speaker # (e.g. Speaker 1, Speaker 2, Speaker 3, etc.).
- You will need to manually tag the unknown speakers.
- No. You can only begin tagging a speaker after the recording has stopped.