Skip to main content

Triggering a VoiceAgent Call from Make

Trigger CloudTalk VoiceAgent calls via Make with HTTP requests. Automate calls, pass data, and customize - no coding needed.

J
Written by Jane Doe
Updated this week

This guide walks you through how to trigger a VoiceAgent call using a Make automation platform and CloudTalk’s API. With just a few configuration steps, you’ll be able to automate outbound calls powered by your VoiceAgent, passing custom data and capturing results - all without writing a single line of code.

User level:

  • Admin


CloudTalk VoiceAgents in Make

By combining the power of Make with CloudTalk VoiceAgent, you can build smart, automated calling workflows that trigger in real time. Whether it's sending reminders, qualifying leads, or collecting feedback, this setup helps you scale outreach, reduce manual effort, and deliver consistent voice experiences. From sales to support, this integration gives your team a powerful way to engage customers through voice with speed and simplicity.

To trigger a VoiceAgent (VA) call, you will need to use an HTTP request in your Make scenario to send request towards Cloudtalk VoiceAgent API.

Prerequisites

Before setting up the HTTP request in Make, make sure you have:

  • CloudTalk API Credentials

    • Go to CloudTalk Dashboard → Account → Settings → API Keys

    • Copy both the API Key ID and API Key Secret

  • VoiceAgent ID

    • VoiceAgent ID can be retrieved from Cloudtalk Dashboard > VoiceAgents

Setup Instructions

To set up the HTTP request in your Make scenario:

  1. In your Make scenario, add a module and select the HTTP Request app.

  2. Select Make a basic auth request action.

  3. Configure authentication

    • Click Add to create new credentials.

    • Use your API Key ID as the Username.

    • Use your API Key Secret as the Password.

    • Give the connection a name and click Create.

  4. For API Endpoint use this URL:
    https://api.cloudtalk.io/v1/voice-agent/calls

  5. For Method select POST

  6. Configure the request format

    • Body type: Raw

    • Content type: application/json

  7. Build the Request content following API documentation

    • Create a JSON payload using the following format:

      {
      "call_number": "{{phone_number}}",
      "voice_agent_id": "{{voice_agent_id}}",
      "call_properties": {
      "system_prompt": {
      "variables": {
      "contact_name": "{{name}}"
      }
      },
      "output": {
      "variables": {
      "custom_field": "value"
      }
      }
      }
      }

    • Replace the variables ({{...}}) with data from earlier steps in your scenario.

    • In call_number, insert a reference to the phone number (e.164) to be called, extracted from previous steps.

    • In voice_agent_id, insert the ID of the VoiceAgent obtained from the CloudTalk Dashboard.

    • If necessary, configure call_properties, where:

      1. Data in call_properties.system_prompt.variables includes request-specific information the VoiceAgent should use during the call (e.g. the contact's name).

      2. Data in call_properties.output.variables defines the values VoiceAgent should pass to the external endpoint after the call, in addition to information extracted during the conversation.

📘 Want to go deeper? Learn how System prompt variables and Output variables shape your VoiceAgent's behavior in this dedicated article!

8. Double-check the configuration and save the module

You’re all set! 🚀 Your setup is complete and ready to start triggering VoiceAgent calls with ease. Let the automation begin!

💡 API Limits to Keep in Mind

To avoid failures or throttling, make sure your Make scenario handles call distribution carefully. Use delays, queues, or scheduling logic where needed, as the standard API Rate limit (60 requests per minute) applies to VoiceAgent endpoint.


If you need any further assistance please do not hesitate to reach out to our Support team, we are always happy to help!

Did this answer your question?