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

# Recommendations

> Guide your users with intelligent question suggestions that appear at the start of a new chat

<Warning>
  **camelAI Legacy Product** — This documentation covers camelAI's embedded analytics offering, which is no longer being actively developed. We are migrating existing customers to the new camelAI platform. For the current product, visit [camelAI](https://camelai.dev).
</Warning>

## Start Recommendations

These are the three recommendation options that appear on the start screen when users starts a new chat.

<img src="https://imagedelivery.net/0Ey8LwpQ4ATeP19F21mqig/2ef487b2-3a2c-4691-8fd7-0893686b6000/public" alt="Show Analysis and Code" height="200" className="rounded-lg" />

<CardGroup cols={2}>
  <Card title="Highlight Common Use Cases" icon="star">
    Showcase the most frequently asked questions in your organization
  </Card>

  <Card title="Showcase Specific Metrics" icon="chart-line">
    Feature important KPIs and metrics users should explore
  </Card>

  <Card title="Guide Toward Valuable Insights" icon="lightbulb">
    Direct users to high-impact analysis opportunities
  </Card>

  <Card title="Demonstrate Capabilities" icon="sparkles">
    Show the range of questions camelAI can answer
  </Card>
</CardGroup>

### Configuration

<Steps>
  <Step title="Provide Starter Prompts">
    Create a list of 1-500 recommended starter prompts
  </Step>

  <Step title="Support Multiple Options">
    If you provide more than 3, camelAI will **randomly select** which 3 to display
  </Step>

  <Step title="Keep It Conversational">
    Recommendations should be conversational and relevant to your data. Keep each recommendation short and sweet.
  </Step>
</Steps>

## In-Chat Recommendations

After each interaction, camelAI automatically generates three follow-up suggestions based on:

* The most recent message exchange
* Your connected data source structure
* Natural next steps in the analysis

<Warning>
  In-chat recommendations are currently AI-generated and cannot be customized. They're designed to be contextually relevant and encourage deeper data exploration.
</Warning>