CallAnalysisTranscriptDTO

One transcript line. The speaker is referenced by name (and optionally role) rather than by a synthetic key: the platform matches on speakerName first, then on speakerRole, and creates a new speaker when neither matches.

speakerNamestringnullable

Name of the speaker of this line. Should match a speakers[].name.

Possible values: <= 255 characters

Example: Aziz Karimov

speakerRoleSpeakerRolenullable

Fallback used to match the speaker when speakerName does not match.

Possible values: [MANAGER, CLIENT, OTHER]

Example: MANAGER

startSecint32nullable

Start offset of the line, in seconds from the beginning of the call.

Possible values: >= 0

Example: 0

endSecint32nullable

End offset of the line, in seconds from the beginning of the call.

Possible values: >= 0

Example: 6

textstringnullable

The spoken text. No length limit.

Example: Assalomu alaykum, Uysot kompaniyasidan Aziz.

CallAnalysisTranscriptDTO

{
  "speakerName": "Aziz Karimov",
  "speakerRole": "MANAGER",
  "startSec": 0,
  "endSec": 6,
  "text": "Assalomu alaykum, Uysot kompaniyasidan Aziz."
}