> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dialflo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Make your first AI-powered voice call with Dialflo in minutes.

Get up and running with **Dialflo** quickly.\
This guide walks you through authenticating, configuring an agent, making your first outbound call, and retrieving call insights.

***

## Get started in three steps

You’ll:

1. Get your Dialflo API Token and authenticate your requests
2. Authenticate using your Dialflo API token
3. Configure an AI agent (one-time setup)
4. Make your first call and retrieve insights

***

### Step 1: Get your Dialflo API token

All Dialflo API requests are authenticated using an API token issued by Dialflo to your organization.

Contact us to get the API token for your Voice calling needs - [abhishek@dialflo.ai](mailto:abhishek@dialflo.ai)

<AccordionGroup>
  <Accordion icon="key" title="How authentication works">
    Each Dialflo customer (organization) is issued an API token during onboarding.
    This token authorizes all API requests made on behalf of your organization.
  </Accordion>

  <Accordion icon="code" title="Authentication headers">
    Include the following headers with every request:

    ```json theme={null}
    {
      "auth_token": "Bearer <YOUR_DIALFLO_API_TOKEN>",
      "Content-Type": "application/json"
    }
    ```

    <Tip>
      Store your API token securely (for example, as an environment variable).
    </Tip>
  </Accordion>
</AccordionGroup>

***

### Step 2: Launch your AI agent (one-time setup)

Before making calls, you need to configure an **AI agent**.

An agent represents the voice identity and phone number used to place outbound calls.

Contact us to get the perfect agent ready for your needs - [abhishek@dialflo.ai](mailto:abhishek@dialflo.ai)

<AccordionGroup>
  <Accordion icon="user-gear" title="Create or configure an agent">
    Agent configuration is a one-time setup and is currently done through the
    Dialflo dashboard or during onboarding with the Dialflo team.

    Once configured, you will receive:

    * An agent phone number **or**
    * An agent name

    You'll use one of these values when initiating calls.

    **API endpoint to create an agent:**

    ```bash theme={null}
        curl --location --globoff '{{ENV_URL}}/agents/' \
          --header 'auth_token: Bearer {{AUTH_TOKEN}}' \
          --header 'Content-Type: application/json' \
          --data '{
            "prompt": {
              "default": "<ENTER_YOUR_AGENT_PROMPT>"
            },
            "agent_name": "<NAME_YOUR_AGENT>",
            "agent_phone_number": "ENTER_PROVIDED_AGENT_NUMBER",
            "voice_id": "ENTER_DESIRED_VOICE_OF_THE",
            "version": 1,
            "welcome_text": {
              "default": "ENTER_GREETING_MESSAGE"
            },
            "tools": null,
            "enabled": true,
            "save_stats": true,
            "server_type": "DEV",
            "call_type": "SELECT_CALL_TYPE",
            "created_at": "<>",
            "updated_at": "<>"
          }'
    ```
  </Accordion>
</AccordionGroup>

***

### Step 3: Make your first AI-powered call

Use the `/make_call` endpoint to initiate an outbound call with personalized context.

```bash theme={null}
curl https://api.dialflo.ai/make_call \
  -H "auth_token: Bearer <YOUR_DIALFLO_API_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
    "from_agent_name": "YOUR_AGENT_NAME",
    "to_phone_number": "<CUSTOMER_PHONE_NUMBER>",
    "extra_config": {
      "call_context": {
        "USER_NAME": "Ankita",
        "JOB_TYPE": "PICKER",
        "CITY_TYPE": "DELHI"
      }
    }
  }'
```

<Tip>
  Store the <code>task\_id</code> returned from the <code>/make\_call</code> request — it is required to retrieve call insights later.
</Tip>

***

### Step 4: Retrieve call details and insights

Once the call is completed, Dialflo generates structured insights and metadata for the conversation.
2025-01-04T15
Use the `task_id` returned in Step 2 to fetch call details.

```bash theme={null}
curl https://api.dialflo.ai/call_details/<TASK_ID> \
  -H "auth_token: Bearer <YOUR_DIALFLO_API_TOKEN>"
```

**Response includes:**

* Call status (completed / failed)
* Duration and timestamps
* Conversation insights and extracted variables
* Recording and transcript references (if enabled)

***

## Next steps

Now that you've made your first call, explore:

* Check out the best practices to cater to your voice powered agentic needs
* Webhook integrations for real-time call events
* Bulk calling capabilities
* Custom workflows and insight extractions