Get Transcript API for AutoTranscribe

Feature Release

This is the announcement for an upcoming ASAPP feature. Your ASAPP account team will provide a target release date and can direct you to more detailed information as needed.

Overview

ASAPP is adding a new endpoint that retrieves the full set of messages for a specified conversation. This expands the delivery use cases for AutoTranscribe, providing a means to get a complete transcript at the end of a conversation on-demand, instead of in real-time during the conversation or in daily batches of conversations.

Use and Impact

This new endpoint is intended to support use cases for full transcripts that require them for end-users or automated services once a conversation ends, rather than on a delayed basis. Potential use cases include:

• Coaching and quality assurance review
• After-call dispositioning workflows
• Agent reference for recent calls
• Text analytics In rare cases where real-time transcript delivery fails, this endpoint also functions as a fallback option to retrieve conversation messages.

How It Works

GET /conversation/v1/conversation/messages takes a single conversation identifier and immediately returns an array of messages that covers the full conversation. Each message includes the transcribed text of the utterance, the role and unique identifier of the sender (agent or customer), and the utterance’s created timestamp. For example:

{
"text": "Hello, I’d like to upgrade my plan to gold.",
"sender": {
  "role": "customer",
  "externalId": "123",
 },
"timestamp": "2022-11-23T12:13:14.555Z"
}

This endpoint should not be used for retrieving individual messages in real-time.

Configuration Transcription settings (e.g. redaction, language) must be configured as part of implementing AutoTranscribe and will be reflected in the messages returned from this endpoint. No further configuration is required.

FAQs

1. Can this endpoint be called before a conversation ends? This endpoint is designed only for after-call retrieval of a transcript. AutoTranscribe offers separate real-time mechanisms for use cases that require conversation utterances during the call. Refer to the AutoTranscribe Deployment Guides for WebSocket and media gateway respectively for more information about real-time transcript delivery.
2. How does AutoTranscribe return conversation messages for multiple conversations? This endpoint only returns transcripts for a single conversation. ASAPP’s File Exporter API provides a batch mechanism for retrieving transcripts for multiple conversations at once. The data source is updated daily and returns utterances for all conversations in the update interval.