Send 1000 Phone Calls

This tutorial will guide you through sending batch phone calls using our API. You'll learn how to set up authorization, create prompts, and handle responses.

Step 1: Setup Your Authorization

To initiate a batch call, you must first authenticate your request. Ensure you have your API key from signing up on the developer portal.

Step 2: Create the Base Prompt

Craft a base prompt that will be common across all calls in the batch. Use placeholders {{curly braces}} for dynamic content.

Example:

'You are calling {{business}} to renew their subscription to {{service}} before it expires on {{date}}.';

Step 3: Define the Call Data

Specify the list of calls in the call_data array. Each call must have a phone_number and can include other properties corresponding to placeholders in your base prompt.

Example:

[
	{
		phone_number: '1234567890',
		business: 'ABC co.',
		service: 'Netflix',
		date: 'September 4th'
	},
	{
		phone_number: '32176540987',
		business: 'XYZ inc.',
		service: 'Window Cleaning',
		date: 'December 20th'
	}
];

Step 4: Additional Configuration

You can customize your batch calls with the following options:

• label: Assign a label to your batch for easy tracking.
• campaign_id: Organize related batches under a campaign.
• test_mode: Set to true for testing with the first call only.
• batch_id: Manually set or auto-generated for tracking.
• Voice and Language Settings: Select voice_id, reduce_latency, and language.
• request_data: Include specific facts for the AI to know during the call.
• webhook: For real-time notifications and transcripts post-call.
• max_duration: Define the maximum length of each call.
• amd: Enable for navigating phone trees or leaving voicemails.
• wait_for_greeting: Control if the AI speaks immediately or waits.

Step 5: Send the API Request

Use the provided JavaScript or Python code snippet to make the API request:

// Headers
const headers = {
	authorization: 'YOUR-API-KEY-HERE'
};

// Data
const data = {
	base_prompt:
		'You are calling {{business}} to renew their subscription to {{service}} before it expires on {{date}}.',
	call_data: [
		{
			phone_number: '1234567890',
			business: 'ABC co.',
			service: 'Netflix',
			date: 'September 4th'
		},
		{
			phone_number: '32176540987',
			business: 'XYZ inc.',
			service: 'Window Cleaning',
			date: 'December 20th'
		}
	],
	label: 'Renewal Reminder - Wednesday Afternoon with female voice',
	voice_id: 0,
	max_duration: 10,
	reduce_latency: true,
	wait_for_greeting: true
};

// API request
const response = await fetch('https://api.example.com/v1/batches', {
	method: 'POST',
	headers,
	body: JSON.stringify(data)
});

Step 6: Handle the Response

After sending the batch request, you'll receive a response with a message and the batch_id. Monitor the progress of your calls and any responses via your specified webhook.

Here's what an example response might look like:

{
	"message": "success",
	"batch_id": "3p$7rQ3p9sT5bzmF-gen-batch"
}

Step 7: Monitoring and Analytics

Track the performance and outcomes of your batch calls through the provided batch_id and campaign analytics. Adjust future batches based on the insights gained.

Best Practices

• Always test with test_mode: true before sending large batches
• Use descriptive labels to organize your batches
• Monitor your webhook for real-time updates
• Set appropriate max_duration limits to control costs
• Use reduce_latency for faster response times when needed