
Dataverse currently stores Teams/voice-channel call transcripts (e.g., in ActivityPointer/PhoneCall/Msdyn_Recording) only as flat text without speaker segmentation;
there’s no built-in entity that holds a diarized version (“who said what”). If you need speaker-level detail, check for any transcript-related entities or attached JSON blobs in your environment,
confirm whether diarization is available as a configuration in your voice-channel or Azure Speech settings, or consider post-processing the audio with a diarization service (e.g., Azure Speech) to enrich the stored transcript.