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

# Tab: Options

> Assign a phone number, set call duration limits, configure voicemail, AI data extraction, and post-call actions.

## Overview

The **Options** tab contains all remaining settings relevant to call operations: phone number assignment, cost protection, call recording, voicemail, and the two key features **AI Data Extraction** and **Post-Call Actions**.

***

## Phone Number

Assign a phone number to the agent – the number from which it will place calls. All numbers you have registered under **Phone Numbers** in SalesFrank are available for selection.

***

## Maximum Call Duration

Set a maximum call duration (e.g., 10 minutes). Calls exceeding this limit are ended automatically. This acts as a **cost safeguard** to prevent individual conversations from running disproportionately long.

***

## Call Recording

Enable or disable recording of calls as audio files. Be aware of the applicable **data privacy regulations** in your region. In many countries, all parties must be informed about recording.

***

## Voicemail Detection

### Enable

Enable voicemail detection so Frank automatically recognizes when an answering machine picks up instead of a real person.

### Voicemail Message

When voicemail detection is active, you can define a **voicemail message**. Frank will automatically speak this message when it detects a voicemail. Lead variables are available here as well for personalization.

***

## AI Data Extraction

This feature lets you define which information Frank should **automatically extract from the transcript** after each call.

### How It Works

After every call, the full transcript is sent to a dedicated AI model. This model's job is to extract or classify the variables you have defined. Extracted values appear in the **lead table** and are also available in **Post-Call Actions**.

### Creating a Variable

For each variable, define:

| Field             | Description                                                                  |
| ----------------- | ---------------------------------------------------------------------------- |
| **Variable name** | A unique identifier (e.g., `email_captured`, `budget`, `interest_product_a`) |
| **Type**          | `Text`, `Boolean` (Yes/No), or `Enum` (selection from predefined values)     |
| **Description**   | Instruction for the AI: what to extract or classify                          |

The **description** is the critical instruction for the AI. Write it clearly and precisely.

**Examples:**

| Variable               | Type                 | Description                                              |
| ---------------------- | -------------------- | -------------------------------------------------------- |
| `email`                | Text                 | The email address the lead mentioned during the call     |
| `appointment_interest` | Boolean              | Did the lead show interest in booking an appointment?    |
| `product_category`     | Enum (`A`, `B`, `C`) | Which product category did the lead express interest in? |
| `budget_available`     | Boolean              | Did the lead indicate that a budget is available?        |

***

## Post-Call Actions

After each conversation, SalesFrank can automatically trigger one or more actions. There are two types:

### Email Notification

Configure one or more email notifications to be sent after specific calls.

**Settings per notification:**

| Field         | Description                                              |
| ------------- | -------------------------------------------------------- |
| **Name**      | Label for the notification (internal use)                |
| **Subject**   | Email subject line                                       |
| **Recipient** | Email address (internal recipients only – not for leads) |
| **Condition** | When should the email be sent?                           |
| **Body**      | Free-form email text with variables                      |

<Warning>
  Email notifications are exclusively for **internal recipients** (e.g., yourself or your team). They cannot be used to send emails to leads.
</Warning>

**Condition types:**

*Predefined conditions* – based on call outcome:

* Always
* Outcome is `INTERESTED`
* Outcome is `NOT_INTERESTED`
* Outcome is `CALLBACK_REQUESTED`
* Outcome is `MEETING_BOOKED`
* Outcome is `NO_ANSWER`
* Outcome is `CALL_TRANSFER`

*Custom AI condition* – write your own condition prompt in plain language. The AI evaluates after each call whether the condition is met.

**Available variables in the email body:**

* Lead standard fields (name, phone, email, company)
* All custom lead variables
* Call outcome
* Transcript summary
* Link to the call
* Call duration
* All AI data extraction variables

***

### API Request (Webhook)

SalesFrank sends an **HTTP POST request** to a URL you specify after every call. The exact structure of the request is visible in the dashboard.

This lets you integrate SalesFrank seamlessly into your own processes – for example with:

* [n8n](https://n8n.io)
* [Make.com](https://make.com)
* [Zapier](https://zapier.com)
* Your own CRM or backend