> For the complete documentation index, see [llms.txt](https://help.verkada.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.verkada.com/intercom/configuration/create-and-manage-ai-receivers.md).

# Create and Manage AI Receivers

Learn how to create, configure, and add AI Receivers to call flows in Verkada Command.

An AI Receiver answers intercom calls with a spoken welcome message, then uses a skill to route the caller or answer questions. Create the receiver on the **Receivers** page, configure its skill, test it in Command, and add it to a call flow.

{% hint style="danger" %}
Your role determines whether you can view, create, or edit AI Receivers. See [AI Receiver permissions by role](#ai-receiver-permissions-by-role) for more information.&#x20;
{% endhint %}

***

### Create an AI Receiver

{% stepper %}
{% step %}
**In Verkada Command, go to All Products > Intercom.**
{% endstep %}

{% step %}
**In the left navigation, select Receivers.**
{% endstep %}

{% step %}
**Click Add AI Receiver.**

Command generates a name for the new receiver. Keep it or replace it with a custom name.
{% endstep %}

{% step %}
**Enter a Welcome Message.**

This is the greeting the AI Receiver says when it answers a call. The AI Receiver repeats this message verbatim at the start of every call.

{% hint style="info" %}
We recommend including a note in the welcome message that tells the caller they are speaking with AI, not a person.
{% endhint %}
{% endstep %}

{% step %}
**Choose a Starting Language from the searchable dropdown of supported languages.**
{% endstep %}

{% step %}
**Choose a Speaking Voice.**
{% endstep %}

{% step %}
**Select a Skill and configure it.**

See Configure your AI Receiver's skill.
{% endstep %}

{% step %}
**Click** [**Test Receiver**](#test-and-save-your-ai-receiver) **and confirm the receiver responds as expected.**
{% endstep %}

{% step %}
**Click Save.**

After you save, the AI Receiver appears in the AI Receivers table. Open it again to edit its configuration or delete it.
{% endstep %}
{% endstepper %}

***

### Configure your AI Receiver's skill

A skill is the action your AI Receiver can perform. Command shows different configuration options based on the skill you select. AI Receivers support two skills:

* **Voice Directory**—Routes callers to the right person or tenant
* **Info Desk**—Answers visitor questions using documents and instructions you provide

After you complete the basic setup, choose and configure the skill.

{% tabs %}
{% tab title="Voice Directory" %}
The Voice Directory skill routes callers to the right person or tenant. A few behaviors apply when you use it across multiple intercoms:

* Voice Directory is available on the TD63 and TK74 Intercom models.
* Command automatically migrated existing Voice Directories to the AI Receivers framework.

{% hint style="warning" %}
For the Voice Directory skill, you can assign the AI Receiver only to the **Call Button** on TD63 when Keypad Directory is enabled. See [Set up TD63 Keypad Directory](/intercom/configuration/wire-a-video-intercom/set-up-td63-keypad-directory.md) for more information.&#x20;
{% endhint %}
{% endtab %}

{% tab title="Info Desk" %}
The Info Desk skill lets your AI Receiver answer visitor questions using information you provide.

{% stepper %}
{% step %}
**Open your AI Receiver's configuration and select the Info Desk skill.**
{% endstep %}

{% step %}
**Under File Upload, add reference documents for the agent to use.**

Supported file types: PDF, CSV, and TXT.
{% endstep %}

{% step %}
**(Optional) Add instructions in the Instruction box.**

This free-text field (up to 5,000 characters) describes how the agent should handle specific situations at your site.

{% hint style="info" %}
You don't need to add extensive guardrails in the Instruction box. Verkada includes a system prompt that blocks abusive or inappropriate content.
{% endhint %}
{% endstep %}
{% endstepper %}
{% endtab %}
{% endtabs %}

***

### Test and save your AI Receiver

Test the AI Receiver in Command before you add it to a call flow.

{% stepper %}
{% step %}
**After you complete the general and skill configuration, click Test Receiver.**
{% endstep %}

{% step %}
**Have a live voice conversation with the receiver to confirm that it responds as you expect.**
{% endstep %}

{% step %}
**If something isn't right, return to the skill configuration, make adjustments, and test again.**
{% endstep %}
{% endstepper %}

#### Save your AI Receiver

When you save, Command starts training the agent with the configuration and documents you provided. You must save again after any further changes.

{% hint style="warning" %}
If this AI Receiver is already in use, clicking **Save** applies your changes to every intercom that currently uses it.
{% endhint %}

***

### Add an AI Receiver to a call flow

You add an AI Receiver to a call flow the same way you add any other receiver.

{% stepper %}
{% step %}
**In Verkada Command, go to All Products > Intercom.**
{% endstep %}

{% step %}
**In the left navigation, select Call Flows.**
{% endstep %}

{% step %}
**Click Add AI Receiver.**
{% endstep %}

{% step %}
**Search for the AI Receiver by name or by type, then add it.**
{% endstep %}
{% endstepper %}

Keep these rules in mind:

* An AI Receiver must be the only receiver in its call flow step. You can't pair it with a human receiver in the same step.
* You can only add an AI Receiver as the final step in a call flow.
* For the Voice Directory skill, you can only assign it to the call button on TD63 while Keypad Directory is enabled.

***

### View AI Receivers in events

Command treats AI Receivers like other receivers in Events and Global Events:

* Calls answered by an AI Receiver count as answered calls.
* The Events timeline shows when an AI Receiver answered a call.
* You can play back the recording and view the transcript of any AI Receiver conversation.
* Filter by **Receiver Type** to show only AI Receivers, or filter by **Receiver** to isolate a specific AI Receiver by name.

***

### AI Receiver permissions by role

| Role        | Permissions                                                                                                                                                                               |
| ----------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Org Admin   | Create and view all AI Receivers in the organization                                                                                                                                      |
| Site Admin  | <p></p><ul><li>Create and view all AI Receivers in the organization</li><li>Update AI Receivers that affect only their site</li><li>Add AI Receivers to their site's call flows</li></ul> |
| Site Viewer | View all AI Receivers in use in their site                                                                                                                                                |

See [Roles and Permissions for Intercom](/intercom/getting-started/roles-and-permissions-for-intercom.md) for more information.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://help.verkada.com/intercom/configuration/create-and-manage-ai-receivers.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
