Cross-channel support

Conversation Intelligence integrates with Conversation Orchestrator. Conversations represent communications between multiple participants that occur across one or more Twilio channels.

Conversation Orchestrator ingests and normalizes these communications into a Conversation and can automatically activate your intelligence configurations.

To use Conversation Intelligence, you must associate your intelligence configuration with one or more conversation configurations.

The following describes how conversation and intelligence configurations interact:

1. The conversation configuration activates your intelligence configuration.
2. The configuration of your conversation configuration, such as which channels send data and how to create conversations, determines which interactions your intelligence configuration analyzes.
3. When conversation events occur (messages, recordings, or transcript segments), the conversation configuration does the following:
   • Detects the event
   • Applies your intelligence configuration rules
   • Triggers the corresponding language operators

Learn more about Ingestion modes.

For per-channel details, see Channels.

To group a customer's interactions across different channels and addresses into a single conversation, set conversationGroupingType to GROUP_BY_PROFILE on your conversation configuration. Without it, calls and messages from the same customer may land in separate conversations and won't share context.

For full details on grouping types and behavior, see Conversation grouping.

Voice has multiple ingestion paths depending on whether you need real-time transcription, post-call transcription, or both.

Define capture rules in your conversation configuration. No changes to your TwiML or voice application are required. Conversation Orchestrator picks up matching inbound and outbound calls automatically.

Make your capture rules bidirectional to catch both inbound and outbound traffic. For capture rule syntax and address formats, see Channels.

To pass real-time transcription data to Conversation Orchestrator and Conversation Intelligence, include conversationConfiguration or conversationId as optional parameters in one of the following:

• Real-time Transcriptions resource
• TwiML™: <ConversationRelay>
• TwiML™: <Transcription>

For more information on how the parameters work and which TwiML verbs to use, see Active TwiML for voice.

Chained configuration automates the workflow from audio capture to AI-driven insights. By linking specific API resources, you can eliminate manual handoffs and automatically trigger Conversation Orchestrator and Conversation Intelligence for post-call batch transcription.

Learn how to set up chained configuration for voice.

You can use both passive and active ingestion for messaging channels. Conversation Orchestrator automatically normalizes inbound and outbound messages into a single conversation.

Define capture rules in your conversation configuration. Conversation Orchestrator picks up matching inbound and outbound messages automatically.

Make your capture rules bidirectional to catch both inbound and outbound traffic. For capture rule syntax and address formats, see Channels.

Create conversations and messaging participants directly through the Conversation Orchestrator API or TwiML. Use active ingestion for outbound campaigns or agent desktop flows where you know the participants up front. For a full walkthrough, see Create conversations programmatically.

Chat is available only through the Conversations API (classic) bridge.

• Set up chained configuration for voice
• Conversation Orchestrator
• Conversation Orchestrator: Ingestion mode
• Conversation Orchestrator: Channels
• Create conversations programmatically