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

# Advanced Chat Techniques

> Master BuildBetter's Chat with advanced prompting and context management

## Advanced Chat Techniques

### Refining AI Outputs

<AccordionGroup>
  <Accordion title="Formatting Commands">
    * "Summarize this in bullet points"
    * "Make this more technical"
    * "Convert to a table format"
    * "Highlight key action items"
  </Accordion>

  <Accordion title="Context References">
    * Reference calls by name: "Use insights from the Q2 Planning call"
    * Filter by signal type: "Focus on bug reports from last month"
    * Combine multiple sources: "Compare feedback across these three customer calls"
  </Accordion>

  <Accordion title="Iterative Refinement">
    * Start broad, then narrow down
    * Ask for specific examples or evidence
    * Request alternative perspectives
    * Polish final output with formatting preferences
  </Accordion>
</AccordionGroup>

## Advanced Prompting Techniques

### Context Management

<Steps>
  <Step title="Set the Scope" stepNumber={1} icon="crop">
    * Specify time periods: "Focus on Q2 2024 calls only"
    * Filter by topic: "Look at feature requests related to the API"
    * Include/exclude sources: "Use customer calls but ignore internal meetings"
  </Step>

  <Step title="Add Constraints" stepNumber={2} icon="filter">
    * Set output format: "Generate a markdown table"
    * Specify length: "Summarize in 3-5 bullet points"
    * Define audience: "Write for technical stakeholders"
  </Step>

  <Step title="Request Analysis" stepNumber={3} icon="chart-line">
    * Ask for patterns: "What common themes emerge?"
    * Compare data: "How does this compare to last quarter?"
    * Seek insights: "What unexpected findings stand out?"
  </Step>
</Steps>

## Document Generation

### Template-Based Generation

<CardGroup cols={2}>
  <Card title="PRD Template" icon="file-lines">
    * Problem statement
    * User requirements
    * Technical constraints
    * Success metrics
  </Card>

  <Card title="Meeting Summary" icon="clipboard">
    * Key decisions
    * Action items
    * Follow-up tasks
    * Timeline updates
  </Card>
</CardGroup>

### Advanced Document Features

* Multi-call analysis
* Custom formatting
* Embedded references
* Version tracking

## Expert Tips

### Maximizing AI Understanding

1. Use clear, specific language
2. Break complex requests into steps
3. Provide examples when needed
4. Verify understanding with follow-ups

### Complex Analysis Patterns

<AccordionGroup>
  <Accordion title="Trend Analysis">
    * Track feature requests over time
    * Monitor sentiment changes
    * Identify emerging patterns
  </Accordion>

  <Accordion title="Comparative Analysis">
    * Compare different customer segments
    * Contrast before/after changes
    * Evaluate competing solutions
  </Accordion>

  <Accordion title="Impact Assessment">
    * Measure feature adoption
    * Track problem resolution
    * Gauge customer satisfaction
  </Accordion>
</AccordionGroup>

<Warning>
  Complex queries may take longer to process. Consider breaking them into smaller, focused requests for optimal results.
</Warning>

## Troubleshooting Advanced Usage

### Common Challenges

* Too much context: Try focusing on specific aspects
* Unclear results: Refine prompts with examples
* Missing data: Verify source availability
* Performance issues: Optimize query complexity

<Note>
  Remember that Chat's AI improves with clearer context and iterative refinement. Don't hesitate to break complex tasks into smaller steps.
</Note>