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

# Types of Agents

> Types of agents guide

## Types of Knowledge Agents

Aivah offers different types of Knowledge Agents, each optimized for specific use cases and interaction patterns. Understanding the differences helps you choose the right configuration for your needs.

## Overview of Agent Types

<CardGroup cols={2}>
  <Card title="Standard Knowledge Agent" icon="book-open">
    General-purpose agents for comprehensive knowledge bases with multiple content sources
  </Card>

  <Card title="Presentation Agent" icon="presentation">
    Specialized agents for single-document presentations and slide decks
  </Card>

  <Card title="Real-Time Agent" icon="bolt">
    Live, conversational agents optimized for instant interactions
  </Card>

  <Card title="Multi-Modal Agent" icon="image">
    Advanced agents capable of handling text, images, and multimedia content
  </Card>
</CardGroup>

## Standard Knowledge Agent

The most versatile type of Knowledge Agent, designed for comprehensive information delivery across multiple content sources.

### Key Features

* **Multiple content sources** - Combine text, URLs, and file uploads
* **Comprehensive knowledge base** - Handle diverse information types
* **Flexible interaction modes** - Support both casual and detailed queries
* **Scalable content management** - Easy to expand and modify knowledge base

### Best Use Cases

<AccordionGroup>
  <Accordion title="Customer Support">
    * FAQ databases
    * Product documentation
    * Troubleshooting guides
    * Policy and procedure information
  </Accordion>

  <Accordion title="Internal Knowledge">
    * Employee handbooks
    * Company procedures
    * Training materials
    * Department-specific information
  </Accordion>

  <Accordion title="Educational Content">
    * Course materials
    * Research databases
    * Reference libraries
    * Academic resources
  </Accordion>
</AccordionGroup>

### Configuration Options

<Tabs>
  <Tab title="Content Sources">
    **Multiple Input Types**

    * Direct text entry for specific information
    * URL integration for web content
    * File uploads for documents and PDFs
    * Mix and match sources as needed
  </Tab>

  <Tab title="Response Style">
    **Flexible Interaction**

    * Detailed explanations for complex topics
    * Quick answers for simple queries
    * Context-aware responses
    * Progressive information disclosure
  </Tab>

  <Tab title="Management">
    **Easy Maintenance**

    * Add, edit, and remove content sources
    * Bulk content management operations
    * Export and backup capabilities
    * Version control for content updates
  </Tab>
</Tabs>

## Presentation Agent

Specialized agents designed for interactive presentation experiences with single-document focus.

### Key Features

* **Single-document optimization** - Focused on one presentation or slide deck
* **Interactive navigation** - Users can explore content conversationally
* **Context-aware responses** - Understands slide relationships and flow
* **Visual content integration** - References charts, graphs, and slide elements

### Best Use Cases

<AccordionGroup>
  <Accordion title="Sales and Marketing">
    * Product pitch decks
    * Marketing presentations
    * Proposal documents
    * Campaign materials
  </Accordion>

  <Accordion title="Training and Education">
    * Corporate training decks
    * Educational presentations
    * Workshop materials
    * Onboarding content
  </Accordion>

  <Accordion title="Executive Communications">
    * Strategy presentations
    * Board meeting materials
    * Quarterly reviews
    * Company announcements
  </Accordion>
</AccordionGroup>

### Configuration Details

<Steps>
  <Step title="Document Requirements">
    * PDF format only
    * Single document at a time
    * Selectable text (not image-based)
    * Clear slide structure and numbering
  </Step>

  <Step title="Interaction Features">
    * Slide-specific questions and answers
    * Cross-slide information linking
    * Content summarization capabilities
    * Navigation by topic or slide number
  </Step>

  <Step title="Management Constraints">
    * Replace entire document to update
    * No additional content sources
    * Document-centric content organization
    * Simplified knowledge management
  </Step>
</Steps>

## Real-Time Agent

Optimized for live, instant interactions with minimal response delay.

### Key Features

* **Instant responses** - Sub-second reply times for most queries
* **Live conversations** - Natural, flowing dialogue experiences
* **Context retention** - Remembers conversation history within sessions
* **Voice compatibility** - Works with speech-to-text and text-to-speech

### Best Use Cases

<AccordionGroup>
  <Accordion title="Customer Service">
    * Live chat support
    * Instant FAQ responses
    * Real-time troubleshooting
    * Immediate assistance requests
  </Accordion>

  <Accordion title="Interactive Applications">
    * Voice assistants
    * Chatbot integrations
    * Real-time Q\&A sessions
    * Live demonstrations
  </Accordion>

  <Accordion title="Internal Tools">
    * Employee self-service
    * IT help desk automation
    * Real-time information lookup
    * Quick decision support
  </Accordion>
</AccordionGroup>

### Performance Considerations

<CardGroup cols={2}>
  <Card title="Response Speed" icon="clock">
    Optimized for sub-second response times with efficient content retrieval
  </Card>

  <Card title="Concurrent Usage" icon="users">
    Handles multiple simultaneous conversations without performance degradation
  </Card>

  <Card title="Context Management" icon="memory">
    Maintains conversation context while optimizing for speed
  </Card>

  <Card title="Integration Ready" icon="plug">
    API-first design for seamless application integration
  </Card>
</CardGroup>

## Multi-Modal Agent

Advanced agents capable of processing and responding to multiple content types including text, images, and multimedia.

### Key Features

* **Image processing** - Analyze and respond to uploaded images
* **Document understanding** - Process complex documents with visual elements
* **Mixed media responses** - Provide text, visual, and interactive responses
* **Advanced content types** - Handle charts, diagrams, screenshots, and more

### Best Use Cases

<AccordionGroup>
  <Accordion title="Technical Support">
    * Screenshot troubleshooting
    * Visual problem diagnosis
    * Diagram explanations
    * Step-by-step visual guides
  </Accordion>

  <Accordion title="Educational Applications">
    * Visual learning materials
    * Image-based questions
    * Multimedia course content
    * Interactive learning experiences
  </Accordion>

  <Accordion title="Creative and Design">
    * Design feedback and analysis
    * Visual content creation
    * Artistic interpretation
    * Creative project assistance
  </Accordion>
</AccordionGroup>

### Technical Requirements

<Steps>
  <Step title="Plan Compatibility">
    Multi-modal features require specific Aivah plan levels. Check your plan for availability.
  </Step>

  <Step title="Content Preparation">
    * High-quality images for best analysis
    * Clear visual elements and text
    * Appropriate file formats and sizes
    * Descriptive alt text when possible
  </Step>

  <Step title="Performance Impact">
    * Slightly longer processing times for complex visuals
    * Higher resource usage for multimedia content
    * Bandwidth considerations for image transfer
    * Storage implications for multimedia knowledge bases
  </Step>
</Steps>

## Choosing the Right Agent Type

### Decision Matrix

<Tabs>
  <Tab title="Content Complexity">
    **Simple to Complex**

    * **Single document** → Presentation Agent
    * **Multiple sources** → Standard Knowledge Agent
    * **Mixed media** → Multi-Modal Agent
    * **Any complexity** → Real-Time Agent (with other features)
  </Tab>

  <Tab title="Interaction Style">
    **Response Requirements**

    * **Instant responses** → Real-Time Agent
    * **Detailed explanations** → Standard Knowledge Agent
    * **Visual interactions** → Multi-Modal Agent
    * **Presentation navigation** → Presentation Agent
  </Tab>

  <Tab title="Use Case Priority">
    **Primary Purpose**

    * **Customer support** → Real-Time or Standard
    * **Training delivery** → Presentation Agent
    * **Knowledge base** → Standard Knowledge Agent
    * **Technical support** → Multi-Modal Agent
  </Tab>
</Tabs>

### Feature Comparison

| Feature                  | Standard | Presentation | Real-Time | Multi-Modal |
| ------------------------ | -------- | ------------ | --------- | ----------- |
| Multiple content sources | ✅        | ❌            | ✅         | ✅           |
| Single document focus    | ❌        | ✅            | Optional  | Optional    |
| Instant responses        | Optional | Optional     | ✅         | Optional    |
| Image processing         | ❌        | ❌            | Optional  | ✅           |
| Voice compatibility      | Optional | Optional     | ✅         | ✅           |
| Context retention        | ✅        | ✅            | ✅         | ✅           |
| API integration          | ✅        | ✅            | ✅         | ✅           |

## Combining Agent Types

### Hybrid Configurations

Many agents can combine multiple features:

<AccordionGroup>
  <Accordion title="Real-Time + Multi-Modal">
    Live agents with image processing capabilities for technical support and interactive assistance.
  </Accordion>

  <Accordion title="Standard + Real-Time">
    Comprehensive knowledge bases with instant response capabilities for customer service applications.
  </Accordion>

  <Accordion title="Presentation + Real-Time">
    Interactive presentation agents with live Q\&A capabilities for training and sales scenarios.
  </Accordion>
</AccordionGroup>

### Implementation Strategy

<Steps>
  <Step title="Start Simple">
    Begin with a Standard Knowledge Agent to establish your knowledge base and understand usage patterns.
  </Step>

  <Step title="Add Features Gradually">
    Enable Real-Time API, Multi-Modal processing, or Presentation mode based on user needs and feedback.
  </Step>

  <Step title="Optimize Performance">
    Monitor usage patterns and optimize agent configuration for your specific use case requirements.
  </Step>

  <Step title="Scale as Needed">
    Expand capabilities and create specialized agents for different departments or use cases.
  </Step>
</Steps>

## Migration Between Agent Types

### Changing Agent Configuration

<Steps>
  <Step title="Assess Current Setup">
    Review your existing agent's content, usage patterns, and performance metrics.
  </Step>

  <Step title="Plan Migration">
    Identify which features to add or modify and potential impact on users.
  </Step>

  <Step title="Test Changes">
    Use a duplicate agent to test new configurations before applying to production.
  </Step>

  <Step title="Implement Gradually">
    Roll out changes during low-usage periods and monitor performance closely.
  </Step>
</Steps>

### Migration Considerations

<Warning>
  Some agent type changes may require rebuilding your knowledge base or reconfiguring integrations. Plan accordingly and consider creating backup agents during transitions.
</Warning>

## Best Practices by Agent Type

<AccordionGroup>
  <Accordion title="Standard Knowledge Agents">
    * Organize content logically with clear categorization
    * Regular content audits and updates
    * Balance comprehensiveness with relevance
    * Test across different query types and complexity levels
  </Accordion>

  <Accordion title="Presentation Agents">
    * Ensure slide content is text-heavy and well-structured
    * Use clear, descriptive slide titles and numbering
    * Test navigation with typical user questions
    * Optimize presentation for conversational interaction
  </Accordion>

  <Accordion title="Real-Time Agents">
    * Prioritize response speed optimization
    * Implement proper error handling and fallbacks
    * Monitor performance under load
    * Plan for scalability and concurrent users
  </Accordion>

  <Accordion title="Multi-Modal Agents">
    * Use high-quality, clear images and visuals
    * Provide context for complex visual content
    * Test with various image types and qualities
    * Balance visual processing with response speed
  </Accordion>
</AccordionGroup>

Ready to create your first agent? Start with our [Agent Creation Guide](/knowledge-agent/create-agent) or learn more about [adding content](/knowledge-agent/add-content) to your knowledge base.