Call Insights Event Stream
Learn how to use Call Insights Event Streams to receive real-time webhooks for call summaries, metrics, and quality events. This feature provides deep visibility into the lifecycle and performance of your Twilio Voice calls through asynchronous event notifications.
You can use this guide for notifications, self-service automation, inbound or outbound contact centers, AI agents, call tracking, PSTN connectivity, and AI/ML transcription.
See Related reference documentation to learn more about the reference page elements used in this guide.
Voice Insights provides seven Call Insights Events which can be subscribed to via Event Streams. Together with the Conference Insights Event Stream, these form part of the unified Voice Insights Event Streams.
| Event | Description |
|---|---|
com.twilio.voice.insights.call-event.gateway | Gateway events are generated by call progress events. Events are typically emitted by Voice Insights within 90 seconds. Gateway events are part of the Call Insights Events. |
com.twilio.voice.insights.call-event.sdk | SDK events are generated by Voice SDK events. Events are typically emitted by Voice Insights within 90 seconds. SDK events are part of the Call Insights Events. |
com.twilio.voice.insights.call-metrics.gateway | Gateway metrics are generated every ten seconds by sensors in Twilio's media gateways. Metrics are typically emitted by Voice Insights within 90 seconds. Gateway metrics are part of the Call Metrics. |
com.twilio.voice.insights.call-metrics.sdk | SDK metrics are emitted every ten seconds by sensors in Twilio's Voice SDKs. Metrics are typically emitted by Voice Insights within 90 seconds. SDK metrics are part of the Call Metrics. |
com.twilio.voice.insights.call-summary.complete | The Call Summary provides an overview of metadata and quality metrics for a single call. A completed call summary may take up to thirty minutes to generate following the end of the call. |
com.twilio.voice.insights.call-summary.partial | A partial summary contains a partial set of the data that comprises the Call Summary. Multiple partial summary events will occur as the Call Summary is assembled. The first partial summary will be available within ten minutes of a call ending. |
com.twilio.voice.insights.call-summary.predicted-complete | The completed Call Summary event is based on a thirty minute timer. Twilio uses ML to predict when a Call Summary is actually completed and emits the predicted-complete event based on the analysis of that model. The predicted complete event is typically available within ten minutes of a call ending. |
This guide covers a feature that can support the following use cases:
Use event streams to monitor the delivery and quality of automated alerts. You can track if emergency notifications or appointment reminders are reaching your customers with high audio quality.
To learn more advanced features that you can use with voice notifications, see Voice notifications.
Monitor user interactions within IVRs by tracking gateway events. This helps you identify points where users might struggle with automated menus.
To learn more advanced features that you can use with self-service automation, see Voice self-service automation.
Provide supervisors with real-time data on call quality and connection status. This ensures that agents and customers have clear communication during support interactions.
To learn more advanced features that you can use with inbound contact centers, see Voice inbound call center.
Track metrics for high-volume outbound dialing to ensure your sales or support teams are maintaining optimal connectivity.
To learn more advanced features that you can use with outbound contact centers, see Voice outbound call center.
Integrate call events with your AI workflows to trigger post-call analysis once a summary event is emitted.
To learn more advanced features that you can use with AI agents, see AI agents.
Use call completion and summary events to programmatically trigger the retrieval of transcriptions for machine learning model training.
To learn more advanced features that you can use with AI or ML transcription, see Voice AI/ML transcription.
Explore the following guides to build on what you've learned in this guide:
- Create conference calls: Host multi-party calls and use Conference Insights to monitor them.
- Gather user input via keypad (DTMF Tones): Build interactive menus that generate gateway events.
- How to capture your first payment using
<Pay>: Integrate secure payment processing into your voice workflows.