Real-time monitoring
Real-time monitoring
Real-time monitoring
Real-time monitoring enables live observation of agent conversations via WebSocket and remote control of active calls. This feature provides real-time visibility into conversation events and allows intervention through control commands.
Monitoring sessions stream conversation events in real-time, including transcripts, agent responses, and corrections. You can also send control commands to end calls, transfer to phone numbers, or enable human takeover during active chat conversations.
Connect to a live conversation using the monitoring endpoint:
Replace {conversation_id} with the ID of the conversation you want to monitor.
Authentication requires:
ElevenLabs Agents Write scopeEDITOR access to the agent’s workspacexi-api-key headerBefore monitoring conversations, enable the feature in your agent’s settings:
Choose which events you want to monitor. See Client Events for a full list of available events.
The following events cannot be monitored: VAD scores, turn probability metrics, and pings.
The conversation must be active before you can connect to monitor it. You cannot monitor a conversation before it begins.
Send JSON commands through the WebSocket to control the conversation:
Terminate the active conversation immediately.
Transfer the call to a specified phone number.
transfer_to_number system tool must already be configured in the agent.Inject context or instructions into the active conversation so the agent can use the new information in its responses.
Switch from AI agent to human operator mode for chat conversations.
Send a message to the user as a human operator in chat conversations.
Return control from human operator back to the AI agent.
Real-time monitoring enables several operational scenarios:
Monitor agent conversations in real-time to ensure quality standards and identify training opportunities.
Detect conversations requiring human intervention and seamlessly take over from the AI agent.
Build real-time monitoring dashboards that aggregate conversation metrics and performance indicators.
Supervise multiple agent conversations simultaneously and intervene when necessary.
Implement automated systems that analyze conversation content and trigger actions based on specific conditions.
Use live conversations as training material and provide real-time feedback to improve agent performance.
Monitoring events are sent asynchronously to the conversation and may not arrive in the same order as the core conversation events. When processing events, do not rely on event order to reconstruct exact conversation timing.
The monitoring endpoint streams only text events and metadata. Raw audio data is not included in monitoring events.
Only approximately the last 100 events are cached and available when connecting to an active conversation. Earlier events cannot be retrieved.
VAD scores, turn probability metrics, and ping events cannot be monitored when custom event selection is enabled.
You must connect after the conversation has started. The monitoring endpoint cannot be used before conversation initiation.
API keys must have ElevenLabs Agents Write scope, and you must have EDITOR workspace access
to monitor conversations.
Receive conversation data and analysis after calls complete.
Configure success evaluation and data collection for conversations.
Understand events received during conversational applications.
Learn about the WebSocket API for real-time conversations.