PRODUCTION SCENARIO
A contact center needs, for each call recording, a transcript that labels who is the agent and who is the customer, plus a summary, the call sentiment, and the topics mentioned.
Which Foundry capability delivers all of this in one analysis?
Answering here is anonymous. Nothing is saved unless you sign in.
Show answer and explanation
Answer: The prebuilt-callCenter analyzer in Content Understanding
The prebuilt-callCenter analyzer returns conversation transcripts with speaker role detection, a call summary, call sentiment, and the topics, companies, and people mentioned. Azure Language sentiment analysis takes written text, not audio. A custom analyzer on the prebuilt-document base cannot accept audio, since a custom analyzer inherits one base analyzer per modality. Batch transcription with diarization separates speakers but produces no summary or sentiment.