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

# Create Your First AI Agent | Beenos Solutions Guide

> Build a new AI agent in Beenos Solutions, connect it to a knowledge base, and configure its name, welcome message, and fallback response.

Creating an AI agent in Beenos Solutions takes just a few minutes. Once you connect an agent to one or more knowledge bases, it is ready to answer user questions drawn directly from your content. This guide walks you through every step of the creation process, from naming your agent to writing the messages your users will see first.

<Steps>
  <Step title="Navigate to the Agents section">
    Log in to your Beenos Solutions dashboard and select **Agents** from the left-hand navigation menu. This is where all of your existing agents are listed and where you manage their settings.
  </Step>

  <Step title="Create a new agent">
    Click the **New Agent** button in the top-right corner. In the dialog that appears, enter a **Name** for your agent — this is what users will see in the chat interface — and an optional internal **Description** to help your team understand the agent's purpose.
  </Step>

  <Step title="Connect a knowledge base">
    Under **Connected Knowledge Bases**, select one or more knowledge bases from the dropdown list. The agent will search all connected knowledge bases when generating responses. You can return to this setting at any time to add or remove sources.
  </Step>

  <Step title="Set your messages">
    Write a **Welcome Message** — the first message users see when they open the chat. Then write a **Fallback Message** — what the agent says when it cannot find relevant content to answer a question. Keep the fallback message helpful and direct users toward an alternative support channel if appropriate.
  </Step>

  <Step title="Create your agent">
    Click **Create Agent**. Your new agent appears in the Agents list and is ready to configure and test. No deployment step is needed at this stage — deployment options become available after creation.
  </Step>
</Steps>

## Agent settings overview

Each agent has a set of configurable fields that control its identity and behavior. Here is what each setting does:

| Setting                       | Description                                                                                                                                                 |
| ----------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Name**                      | The display name shown to users in the chat interface and on the shareable agent page.                                                                      |
| **Description**               | An internal note visible only to your team. Use it to document the agent's intended audience or scope.                                                      |
| **Connected knowledge bases** | The knowledge bases the agent searches when generating answers. You can connect multiple bases to broaden coverage.                                         |
| **Welcome message**           | The opening message displayed to users when they start a new conversation. Use it to set expectations about what the agent can help with.                   |
| **Fallback message**          | Shown when the agent cannot find content relevant enough to answer the user's question confidently. A good fallback message guides users toward next steps. |
| **Language**                  | Set to **Auto-detect** to let the agent match the user's language, or fix it to a specific language if your knowledge base is monolingual.                  |

<Tip>
  Before you deploy your agent publicly, use the **Chat Playground** tab on the agent's detail page to test it with real questions. The playground lets you verify that the agent retrieves the right content and that your welcome and fallback messages read naturally.
</Tip>