Diarization
WHAT IT MEANS
Working out who is speaking when, e.g. separating the customer from the agent, or noticing a second person in the room.
Real example
The call transcript is labelled “Agent:” and “Customer:” correctly.
What’s normal
It’s important for call analytics and QA. Less important for live bots that use separate audio channels.
The India angle
Background voices, such as family members or shop noise, are common and can confuse the bot.
Questions to ask your vendor
Do you record each side on a separate channel?
How do you handle background voices?
Is diarization included in transcripts?
Companies that do this
Other terms in Hearing (Speech-to-Text)
Supported by
