Get messages

Get a transcript of all messages in a conversation.

This request provides the following optional features:

• Redaction of detected entities, such as PII, PCI, and PHI data. Use the redact query parameter to enable redaction. When enabled, all detected entities are redacted by default. You can exclude individual entities from redaction using the exclude query parameter.

• Sentiment analysis of sentences in a conversation. Use the sentiment query parameter to enable sentiment analysis. When enabled, each message returned in the response includes the sentiment polarity of the sentence.

• Verbose messages. Use the verbose query parameter to enable verbose message responses. When enabled, each response includes an additional collection of objects, one for each word in the corresponding sentence. When enabled, it generates a new words array, which includes:

  • word - The word being analyzed.
  • startTime - The start time of the word.
  • endTime - The end time of the word.
  • speakerTag - The speaker of the word.
  • score - The contextual score of the word, compared to other words in the sentence and based on the structure of the conversation.
  • timeOffset - The location of the word in the message. The value is the number of characters from the start of the message.
  • duration - The total time that the word was spoken.

Returns a messages object with optional redaction, sentiment analysis, and word-level verbosity.

For a formatted transcript, try Create formatted transcript or Create Summary UI.