Skip to main content

Overview

The Incoming Call Node triggers a workflow automatically when a customer calls your voice agent. This enables you to process call data, perform post-call actions, and integrate with other systems based on inbound customer interactions.

Use Cases

  • Post-Call Processing: Automatically send follow-up emails or SMS after calls
  • CRM Integration: Update customer records with call details
  • Ticket Creation: Generate support tickets based on call content
  • Analytics: Track and analyze inbound call patterns
  • Lead Qualification: Score and route leads based on call outcomes
  • Callback Scheduling: Schedule follow-up calls for unresolved issues
  • Notification Systems: Alert teams about important customer calls

How It Works

  1. Customer calls your voice agent’s phone number
  2. Voice agent handles the conversation
  3. When call ends, Incoming Call Node is triggered
  4. Call details and transcript are passed to subsequent workflow blocks
  5. Automated actions are executed based on your workflow configuration

Outputs

The Incoming Call Node provides comprehensive call information that can be used by subsequent blocks:

Bot_Number

  • Phone number on which the customer called the voice agent
  • The inbound line or SIP address
  • Format: +[country code][number]
Example: +91xxxxxxxxxx

Use_Case

  • The voice agent that handled the call
  • References the agent configured in your account
  • Useful for routing different call types
Example: customer_support, sales_inquiry

Call_Time

  • Timestamp of when the call occurred
  • Format: ISO 8601 date format YYYY-MM-DDTHH:MM:SSZ
  • Includes timezone information
Example: 2025-12-10T14:30:00Z

Call_Status

  • Final status of the call
  • Possible values:
    • connected: Call was successfully connected and completed
    • couldn't connect: Call failed to connect
Example: connected

Call_Duration

  • Duration of the call in seconds
  • Excludes ring time and connection delays
  • 0 for failed connections
Example: 185 (3 minutes 5 seconds)

Call_Direction

  • Direction of the call
  • Always inbound for this node
  • Distinguishes from outbound calls
Example: inbound

Customer_Number

  • Customer’s phone number who made the call
  • Format: +[country code][number]
  • Can be used for callback or identification
Example: +91xxxxxxxxxx

Customer_Details

  • Customer information in JSON format
  • May include: name, email, account ID, etc.
  • Populated from your CRM or previous interactions
Example:
{
  "name": "John Doe",
  "email": "[email protected]",
  "account_id": "ACC-12345",
  "customer_type": "premium"
}

Transcript

  • Complete transcript of the call conversation
  • Includes speaker tags (Assistant vs Customer)
  • Formatted text with timestamps
Example:
Assistant: Hello! Thank you for calling. How can I help you today?
Customer: Hi, I need help with my recent order.
Assistant: I'd be happy to help. Can you provide your order number?
Customer: Yes, it's ORD-12345.
...

Workflow Examples

Example 1: Post-Call Email Follow-up

Workflow:
  1. Incoming Call Node (trigger)
  2. AI Query Agent Node (analyze call and generate email)
  3. Send Email Node (send follow-up to customer)
Use Case: Automatically send a summary email after support calls

Example 2: CRM Update

Workflow:
  1. Incoming Call Node (trigger)
  2. Smart Filter Node (check if call was successful)
  3. Webhook Node (update CRM with call details)
Use Case: Keep customer records updated with call history

Example 3: Callback Scheduling

Workflow:
  1. Incoming Call Node (trigger)
  2. AI Query Agent Node (determine if callback needed)
  3. Schedule Callback Node (schedule follow-up call)
Use Case: Automatically schedule callbacks for unresolved issues

Example 4: Support Ticket Creation

Workflow:
  1. Incoming Call Node (trigger)
  2. AI Query Agent Node (extract issue details)
  3. Webhook Node (create ticket in support system)
  4. Send Email Node (notify customer of ticket number)
Use Case: Convert support calls into trackable tickets

Example 5: Lead Scoring

Workflow:
  1. Incoming Call Node (trigger)
  2. AI Query Agent Node (analyze conversation for buying signals)
  3. Smart Filter Node (filter high-quality leads)
  4. Webhook Node (send qualified leads to sales team)
Use Case: Identify and route hot leads from inbound calls

Configuration

Setting Up Inbound Telephony

Before using the Incoming Call Node, ensure you have:
  1. Configured Inbound Trunk: Set up SIP trunk in Telephony Integration
  2. Mapped Voice Agent: Associate your agent with the inbound number
  3. Tested Connection: Make a test call to verify setup
See Telephony Integration for detailed setup instructions.

Adding to Workflow

  1. Create or edit a workflow
  2. Add the Incoming Call Node as the first block
  3. Connect subsequent blocks to process call data
  4. Save and activate the workflow
The workflow will automatically trigger for all incoming calls to the mapped agent.

Processing Call Data

Using Transcript Data

The transcript can be used for:
  • Sentiment Analysis: Determine customer satisfaction
  • Intent Detection: Identify customer needs
  • Information Extraction: Pull out order IDs, dates, etc.
  • Quality Assurance: Review agent performance
  • Training Data: Improve agent responses
Example with AI Query Agent:
Query: Analyze the call transcript and extract:
1. Customer's main issue
2. Whether issue was resolved
3. Customer sentiment (positive/negative/neutral)
4. Any follow-up actions needed

Using Call Metadata

Call metadata (duration, status, time) can be used for:
  • Performance Metrics: Track average call duration
  • Availability Analysis: Identify peak call times
  • Success Rate: Monitor connection success
  • Reporting: Generate call volume reports

Best Practices

Workflow Design

  • Keep It Simple: Start with basic workflows and add complexity gradually
  • Error Handling: Include fallback actions for failed steps
  • Testing: Test workflows with sample calls before going live
  • Monitoring: Regularly review workflow execution logs

Data Processing

  • Validate Data: Check that required fields are present
  • Handle Failures: Account for calls that couldn’t connect
  • Privacy: Respect customer data privacy regulations
  • Retention: Set appropriate data retention policies

Performance

  • Async Processing: Use webhooks for long-running tasks
  • Batch Operations: Group similar actions when possible
  • Timeout Handling: Set appropriate timeouts for external calls
  • Rate Limiting: Respect API rate limits for integrations

Integration

  • Idempotency: Ensure actions can be safely retried
  • Authentication: Secure all external API calls
  • Error Logging: Log errors for troubleshooting
  • Monitoring: Set up alerts for workflow failures

Monitoring and Analytics

Track Workflow Performance

Monitor your incoming call workflows:
  • Execution Count: How many times workflow triggered
  • Success Rate: Percentage of successful executions
  • Average Duration: Time to complete workflow
  • Error Rate: Failed workflow executions

Call Analytics

Analyze incoming call patterns:
  • Call Volume: Calls per hour/day/week
  • Peak Times: When most calls occur
  • Average Duration: Typical call length
  • Connection Rate: Successful vs failed calls

Business Insights

Extract valuable insights:
  • Common Issues: Frequently discussed topics
  • Customer Sentiment: Overall satisfaction trends
  • Agent Performance: Call handling effectiveness
  • Conversion Rates: Leads to customers ratio

Troubleshooting

Workflow Not Triggering

Possible Causes:
  • Inbound trunk not configured correctly
  • Agent not mapped to phone number
  • Workflow not activated
Solutions:
  • Verify telephony integration setup
  • Check agent mapping in Integrations section
  • Ensure workflow is saved and activated

Missing Call Data

Possible Causes:
  • Call disconnected before completion
  • Transcript processing delay
  • Network issues
Solutions:
  • Check call status in Call History
  • Wait a few minutes for transcript processing
  • Review system status page

Workflow Execution Errors

Possible Causes:
  • Invalid configuration in subsequent blocks
  • External API failures
  • Timeout issues
Solutions:
  • Review workflow execution logs
  • Test each block individually
  • Check external service status

Next Steps

Use the AI Query Agent Node to intelligently process call transcripts and make decisions about follow-up actions.
Transcripts may take 1-2 minutes to process after the call ends. Design your workflows to account for this delay if needed.