Skip to content
AI Employee

Voice AI Custom Actions

This article explains how to use Voice AI Custom Actions to send real-time POST webhook requests during live calls. Custom Actions allow Voice AI agents…

5 min read1,093 words8 explained imagesUpdated Mon, 6 Jul at 7:58 AM
View as markdownOpen in ClaudeOpen in ChatGPT
Each image has a one line explanation. Switch to full detail for the step it belongs to, the fields and buttons involved, examples and the whole procedure.
On this page
  1. Key Benefits of Voice AI Custom Actions
  2. Voice AI Conversation Triggers
  3. Webhook Custom Action
  4. Dynamic Parameter Collection
  5. How to Set Up Voice AI Custom Actions
  6. Step 1: Navigate to your Voice AI Agent
  7. Step 2: Add a Custom Action
  8. Step 3: Configure the Custom Action
  9. Voice AI - Real-time Testing
  10. Frequently Asked Questions
  11. Related Articles

This article explains how to use Voice AI Custom Actions to send real-time POST webhook requests during live calls. Custom Actions allow Voice AI agents to interact with external systems mid-conversation, such as looking up data, sending information, or triggering external processes before the call ends.

TABLE OF CONTENTS

What are Voice AI Custom Actions?

Voice AI Custom Actions allow Voice AI agents to send custom POST webhook requests to external APIs during a live conversation. These actions can include webhook URLs, headers, authentication, and dynamic parameters collected from the caller in real time. This helps the agent retrieve or send information based on what the caller says during the call.

Key Benefits of Voice AI Custom Actions

Voice AI Custom Actions help Voice AI agents connect with external systems during a live call. This allows the agent to personalize responses, retrieve information, and complete tasks faster.

  • Send real-time POST webhook requests based on caller intent or conversation cues.
  • Look up or submit information during the call.
    (Example: If a caller asks, “What’s the status of my recent order?”, the AI can call your order management system and retrieve the order status without putting the caller on hold.)
  • Configure webhook URLs, headers, and authentication.
  • Pass dynamic call data, such as phone numbers, email addresses, dates, or order IDs.
  • Test webhook requests and responses before saving the action
  • Reduce manual follow-up by resolving more requests during the call.

Voice AI Conversation Triggers

Conversation triggers define when a Custom Action should run during a call. A trigger can be based on what the caller says, the information they provide, or the intent detected by the Voice AI agent.

Example triggers:

  • When a caller says, “I want to check my appointment.”
  • When the caller provides an email address.
  • When the caller provides a string of digits, such as an order number, booking ID, or account number.
Voice AI Conversation Triggers (image 1 of 8) What this shows When a caller says, “I want to check my appointment.” When the caller provides an email address. What this shows Illustrates the "Voice AI Conversation Triggers" section of "Voice AI Custom Actions". This screenshot appears in the "Voice AI Conversation Triggers" section of "Voice AI Custom Actions". The text alongside this image reads: When a caller says, “I want to check my appointment.” When the caller provides an email address. When the caller provides a string of digits, such as an order number, booking ID, or account number. This part of the guide covers 1 field, listed below. Immediately after, the guide continues: Webhook integration is the core of Voice AI Custom Actions. It allows your Voice AI agent to interact with external systems that support API requests, such as CRMs, scheduling tools, databases, order management systems, and other business…. Image 1 of 8 What this coversWhen a caller says, “I want to check my appointment.” When the caller provides an email address. When the caller provides a string of digits, such as an order number, booking ID, or account number.
Fields in this part of the guide When a caller says, “I want to check my appointment.”
Next stepWebhook integration is the core of Voice AI Custom Actions. It allows your Voice AI agent to interact with external systems that support API requests, such as CRMs, scheduling tools, databases, order management systems, and other business applications.
All 3 steps in this procedure
  1. When a caller says, “I want to check my appointment.”
  2. When the caller provides an email address.
  3. When the caller provides a string of digits, such as an order number, booking ID, or account number.

Webhook Custom Action

Webhook integration is the core of Voice AI Custom Actions. It allows your Voice AI agent to interact with external systems that support API requests, such as CRMs, scheduling tools, databases, order management systems, and other business applications.

Each Custom Action is defined by a POST request, which may include:

  • A webhook endpoint URL
  • Headers, such as API keys or tokens
  • Authentication, such as Bearer token or Basic Auth
  • A request body with dynamic parameters collected during the call

Example: If you are integrating with a CRM, your webhook URL might target an endpoint such as `/api/v1/lookupContact` and include parameters like the caller’s email address or phone number.

Example: If you’re integrating with a CRM like Salesforce, your webhook URL might target an endpoint like /api/v1/lookupContact and include parameters like the contact’s email or phone number.

Webhook Custom Action (image 2 of 8) What this shows Example: If you’re integrating with a CRM like Salesforce, your webhook URL might target an endpoint like /api/v1/lookupContact and… What this shows Illustrates the "Webhook Custom Action" section of "Voice AI Custom Actions". This screenshot appears in the "Webhook Custom Action" section of "Voice AI Custom Actions". The text alongside this image reads: Example: If you’re integrating with a CRM like Salesforce, your webhook URL might target an endpoint like /api/v1/lookupContact and include parameters like the contact’s email or phone number. This part of the guide covers 3 fields, listed below. Immediately after, the guide continues: Voice AI can extract and label relevant information during a live conversation. These values can be assigned to parameters in the webhook request body. This is useful when the AI needs to collect details like an order number, email…. Image 2 of 8 What this coversExample: If you’re integrating with a CRM like Salesforce, your webhook URL might target an endpoint like /api/v1/lookupContact and include parameters like the contact’s email or phone number.
Fields in this part of the guide A webhook endpoint URLHeaders, such as API keys or tokensAuthentication, such as Bearer token or Basic Auth
Next stepVoice AI can extract and label relevant information during a live conversation. These values can be assigned to parameters in the webhook request body. This is useful when the AI needs to collect details like an order number, email address, phone number, or appointment date before sending the webhook request.
Example values
A webhook endpoint URL
https://northgatedental.com

Illustrative values showing the expected format. They are not read from the screenshot.

All 4 steps in this procedure
  1. A webhook endpoint URL
  2. Headers, such as API keys or tokens
  3. Authentication, such as Bearer token or Basic Auth
  4. A request body with dynamic parameters collected during the call

Dynamic Parameter Collection

Voice AI can extract and label relevant information during a live conversation. These values can be assigned to parameters in the webhook request body. This is useful when the AI needs to collect details like an order number, email address, phone number, or appointment date before sending the webhook request.

Supported data types:

  • Text / String
  • Number / Numeric
  • Email
  • Phone Number
  • Date
Dynamic Parameter Collection (image 3 of 8) What this shows Text / String Number / Numeric Email Phone Number Date What this shows Illustrates the "Dynamic Parameter Collection" section of "Voice AI Custom Actions". This screenshot appears in the "Dynamic Parameter Collection" section of "Voice AI Custom Actions". The text alongside this image reads: Text / String Number / Numeric Email Phone Number Date. This part of the guide covers 5 fields, listed below. Immediately after, the guide continues: Voice AI Custom Actions can be configured from the Voice AI Agent interface with just a few steps. Here’s how to get started. Image 3 of 8 What this coversText / String Number / Numeric Email Phone Number Date
Fields in this part of the guide Text / StringNumber / NumericEmailPhone NumberDate
Next stepVoice AI Custom Actions can be configured from the Voice AI Agent interface with just a few steps. Here’s how to get started:
Example values
Email
priya.raman@northgatedental.com
Phone Number
+1 312 847 1928

Illustrative values showing the expected format. They are not read from the screenshot.

All 5 steps in this procedure
  1. Text / String
  2. Number / Numeric
  3. Email
  4. Phone Number
  5. Date

How to Set Up Voice AI Custom Actions

Voice AI Custom Actions can be configured from the Voice AI Agent interface with just a few steps. Here’s how to get started:

Step 1: Navigate to your Voice AI Agent

Navigate to Voice AI > Agent Goals > Switch to Advanced Mode (If not already enabled) > Custom Actions.

Step 1: Navigate to your Voice AI Agent (image 4 of 8) What this shows Navigate to Voice AI > Agent Goals > Switch to Advanced Mode (If not already enabled) > Custom Actions. What this shows Illustrates the "Step 1: Navigate to your Voice AI Agent" section of "Voice AI Custom Actions". This animation appears in the "Step 1: Navigate to your Voice AI Agent" section of "Voice AI Custom Actions". The text alongside this image reads: Navigate to Voice AI > Agent Goals > Switch to Advanced Mode (If not already enabled) > Custom Actions. The navigation path used here is Voice AI > Agent Goals > Switch. Immediately after, the guide continues: In the Setup your Actions section, click Custom Action to open the custom action configuration window. Path: Voice AI > Agent Goals > SwitchImage 4 of 8 Where to goNavigate to Voice AI > Agent Goals > Switch to Advanced Mode (If not already enabled) > Custom Actions. Next stepIn the Setup your Actions section, click Custom Action to open the custom action configuration window.

Step 2: Add a Custom Action

In the Setup your Actions section, click Custom Action to open the custom action configuration window.

Step 2: Add a Custom Action (image 5 of 8) What this shows In the Setup your Actions section, click Custom Action to open the custom action configuration window. What this shows Illustrates the "Step 2: Add a Custom Action" section of "Voice AI Custom Actions". This screenshot appears in the "Step 2: Add a Custom Action" section of "Voice AI Custom Actions". The text alongside this image reads: In the Setup your Actions section, click Custom Action to open the custom action configuration window. Immediately after, the guide continues: Enter the required Custom Action details. Some fields may be optional depending on your webhook setup. Image 5 of 8 What to clickIn the Setup your Actions section, click Custom Action to open the custom action configuration window. Next stepEnter the required Custom Action details. Some fields may be optional depending on your webhook setup.

Step 3: Configure the Custom Action

Enter the required Custom Action details. Some fields may be optional depending on your webhook setup.

  1. Enter a name for the Custom Action.
  2. Set the conversation trigger conditions.
    Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API.
  3. Add your webhook URL.
  4. Confirm that the request method is POST.
  5. Add any required headers.
  6. Enter authentication details if required, such as a Bearer token or Basic Auth.
  7. Define the dynamic parameters that should be collected from the conversation.
  8. Use the Test Webhook feature to validate the setup before saving.
Step 3: Configure the Custom Action (image 6 of 8) What this shows Enter a name for the Custom Action. What this shows Illustrates the "Step 3: Configure the Custom Action" section of "Voice AI Custom Actions". This screenshot appears in the "Step 3: Configure the Custom Action" section of "Voice AI Custom Actions". The text alongside this image reads: Enter a name for the Custom Action. Set the conversation trigger conditions. Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API. Add your webhook URL. Confirm that the request method is…. Immediately after, the guide continues: Before saving a Custom Action, use the built-in Test Webhook tool to validate the request and response. This allows you to pass test data, review the webhook request, and confirm that your external system returns the expected response. Image 6 of 8 What to fill inEnter a name for the Custom Action. Set the conversation trigger conditions. Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API. Add your webhook URL. Confirm that the request method is POST. Add any required headers. Enter authentication details if required, such as a Bearer token or Basic Auth. Define the dynamic parameters that should be collected from the conversation. Use the Test Webhook feature to validate the setup before saving. Next stepBefore saving a Custom Action, use the built-in Test Webhook tool to validate the request and response. This allows you to pass test data, review the webhook request, and confirm that your external system returns the expected response.
All 8 steps in this procedure
  1. Enter a name for the Custom Action.
  2. Set the conversation trigger conditions. Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API.
  3. Add your webhook URL.
  4. Confirm that the request method is POST.
  5. Add any required headers.
  6. Enter authentication details if required, such as a Bearer token or Basic Auth.
  7. Define the dynamic parameters that should be collected from the conversation.
  8. Use the Test Webhook feature to validate the setup before saving.
Step 3: Configure the Custom Action (image 7 of 8) What this shows Enter a name for the Custom Action. What this shows Illustrates the "Step 3: Configure the Custom Action" section of "Voice AI Custom Actions". This screenshot appears in the "Step 3: Configure the Custom Action" section of "Voice AI Custom Actions". The text alongside this image reads: Enter a name for the Custom Action. Set the conversation trigger conditions. Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API. Add your webhook URL. Confirm that the request method is…. Immediately after, the guide continues: Before saving a Custom Action, use the built-in Test Webhook tool to validate the request and response. This allows you to pass test data, review the webhook request, and confirm that your external system returns the expected response. Image 7 of 8 What to fill inEnter a name for the Custom Action. Set the conversation trigger conditions. Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API. Add your webhook URL. Confirm that the request method is POST. Add any required headers. Enter authentication details if required, such as a Bearer token or Basic Auth. Define the dynamic parameters that should be collected from the conversation. Use the Test Webhook feature to validate the setup before saving. Next stepBefore saving a Custom Action, use the built-in Test Webhook tool to validate the request and response. This allows you to pass test data, review the webhook request, and confirm that your external system returns the expected response.
All 8 steps in this procedure
  1. Enter a name for the Custom Action.
  2. Set the conversation trigger conditions. Example: If a caller asks to “check my account balance,” you can set a trigger for that request and define a phone number parameter to pass into your external API.
  3. Add your webhook URL.
  4. Confirm that the request method is POST.
  5. Add any required headers.
  6. Enter authentication details if required, such as a Bearer token or Basic Auth.
  7. Define the dynamic parameters that should be collected from the conversation.
  8. Use the Test Webhook feature to validate the setup before saving.

Voice AI - Real-time Testing

Before saving a Custom Action, use the built-in Test Webhook tool to validate the request and response. This allows you to pass test data, review the webhook request, and confirm that your external system returns the expected response.

You’ll be able to:

  • Review the full request, including headers and body.
  • View the raw response from the webhook, such as 200 OK, 400 Bad Request, or 404 Not Found.
  • Confirm that the returned data can be used by the Voice AI agent during the call.
  • Identify and fix configuration issues before saving.
Voice AI - Real-time Testing (image 8 of 8) What this shows Review the full request, including headers and body. What this shows Illustrates the "Voice AI - Real-time Testing" section of "Voice AI Custom Actions". This screenshot appears in the "Voice AI - Real-time Testing" section of "Voice AI Custom Actions". The text alongside this image reads: Review the full request, including headers and body. View the raw response from the webhook, such as 200 OK, 400 Bad Request, or 404 Not Found. Confirm that the returned data can be used by the Voice AI agent during the call. Identify and fix configuration issues before saving. Immediately after, the guide continues: Q: Can I use GET or other request types? Image 8 of 8 How to confirmReview the full request, including headers and body. View the raw response from the webhook, such as 200 OK, 400 Bad Request, or 404 Not Found. Confirm that the returned data can be used by the Voice AI agent during the call. Identify and fix configuration issues before saving. Next stepQ: Can I use GET or other request types?
All 4 steps in this procedure
  1. Review the full request, including headers and body.
  2. View the raw response from the webhook, such as 200 OK, 400 Bad Request, or 404 Not Found.
  3. Confirm that the returned data can be used by the Voice AI agent during the call.
  4. Identify and fix configuration issues before saving.

Frequently Asked Questions

Q: Can I use GET or other request types?

No, only POST requests are currently supported for Custom Actions.

Q: Where do I access Voice AI Custom Actions?

You can access Custom Actions by going to AI Agents > Voice AI, then creating or editing a Voice AI agent. In the agent setup area, go to Setup your Actions and select Custom Action.

Q: Is authentication supported in webhooks?

Yes, you can use Bearer tokens, Basic Auth, or pass keys in headers.

Q: What if my webhook fails during the call?

The system logs the failure. You can define fallback behavior for cases where no data is returned, the request fails, or the webhook times out.

Q: Can I trigger multiple webhooks during a single call?

Yes. Each Custom Action can be triggered independently based on its own conditions.

Q: Do I need a developer to set up these actions?

Not always. Basic webhook setups may be configured by users who have the external system’s API documentation. More advanced use cases, authentication requirements, or custom payloads may require help from a developer.

Q: What request method is supported?

Custom Actions currently support POST requests.

Was this guide helpful?

Related guides

Overview of Voice AI Agents6 min read · 10 imagesHow to Create Voice AI Agents9 min read · 8 imagesHow to Test Voice AI Agents7 min read · 5 imagesVoice AI - How to Configure Inbound Call Flow4 min read · 9 images