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

Standard Knowledge Agent

General-purpose agents for comprehensive knowledge bases with multiple content sources

Presentation Agent

Specialized agents for single-document presentations and slide decks

Real-Time Agent

Live, conversational agents optimized for instant interactions

Multi-Modal Agent

Advanced agents capable of handling text, images, and multimedia content

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

Customer Support

• FAQ databases
• Product documentation
• Troubleshooting guides
• Policy and procedure information

Internal Knowledge

• Employee handbooks
• Company procedures
• Training materials
• Department-specific information

Educational Content

• Course materials
• Research databases
• Reference libraries
• Academic resources

Configuration Options

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

Response Style

Flexible Interaction

• Detailed explanations for complex topics
• Quick answers for simple queries
• Context-aware responses
• Progressive information disclosure

Management

Easy Maintenance

• Add, edit, and remove content sources
• Bulk content management operations
• Export and backup capabilities
• Version control for content updates

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

Sales and Marketing

• Product pitch decks
• Marketing presentations
• Proposal documents
• Campaign materials

Training and Education

• Corporate training decks
• Educational presentations
• Workshop materials
• Onboarding content

Executive Communications

• Strategy presentations
• Board meeting materials
• Quarterly reviews
• Company announcements

Configuration Details

1. Document Requirements

   • PDF format only
   • Single document at a time
   • Selectable text (not image-based)
   • Clear slide structure and numbering
2. Interaction Features

   • Slide-specific questions and answers
   • Cross-slide information linking
   • Content summarization capabilities
   • Navigation by topic or slide number
3. Management Constraints

   • Replace entire document to update
   • No additional content sources
   • Document-centric content organization
   • Simplified knowledge management

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

Customer Service

• Live chat support
• Instant FAQ responses
• Real-time troubleshooting
• Immediate assistance requests

Interactive Applications

• Voice assistants
• Chatbot integrations
• Real-time Q&A sessions
• Live demonstrations

Internal Tools

• Employee self-service
• IT help desk automation
• Real-time information lookup
• Quick decision support

Performance Considerations

Response Speed

Optimized for sub-second response times with efficient content retrieval

Concurrent Usage

Handles multiple simultaneous conversations without performance degradation

Context Management

Maintains conversation context while optimizing for speed

Integration Ready

API-first design for seamless application integration

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

Technical Support

• Screenshot troubleshooting
• Visual problem diagnosis
• Diagram explanations
• Step-by-step visual guides

Educational Applications

• Visual learning materials
• Image-based questions
• Multimedia course content
• Interactive learning experiences

Creative and Design

• Design feedback and analysis
• Visual content creation
• Artistic interpretation
• Creative project assistance

Technical Requirements

1. Plan Compatibility

   Multi-modal features require specific Aivah plan levels. Check your plan for availability.

2. Content Preparation

   • High-quality images for best analysis
   • Clear visual elements and text
   • Appropriate file formats and sizes
   • Descriptive alt text when possible
3. 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

Choosing the Right Agent Type

Decision Matrix

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)

Interaction Style

Response Requirements

• Instant responses → Real-Time Agent
• Detailed explanations → Standard Knowledge Agent
• Visual interactions → Multi-Modal Agent
• Presentation navigation → Presentation Agent

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

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:

Real-Time + Multi-Modal

Live agents with image processing capabilities for technical support and interactive assistance.

Standard + Real-Time

Comprehensive knowledge bases with instant response capabilities for customer service applications.

Presentation + Real-Time

Interactive presentation agents with live Q&A capabilities for training and sales scenarios.

Implementation Strategy

1. Start Simple

   Begin with a Standard Knowledge Agent to establish your knowledge base and understand usage patterns.

2. Add Features Gradually

   Enable Real-Time API, Multi-Modal processing, or Presentation mode based on user needs and feedback.

3. Optimize Performance

   Monitor usage patterns and optimize agent configuration for your specific use case requirements.

4. Scale as Needed

   Expand capabilities and create specialized agents for different departments or use cases.

Migration Between Agent Types

Changing Agent Configuration

1. Assess Current Setup

   Review your existing agent’s content, usage patterns, and performance metrics.

2. Plan Migration

   Identify which features to add or modify and potential impact on users.

3. Test Changes

   Use a duplicate agent to test new configurations before applying to production.

4. Implement Gradually

   Roll out changes during low-usage periods and monitor performance closely.

Migration Considerations

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

Best Practices by Agent Type

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

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

Real-Time Agents

• Prioritize response speed optimization
• Implement proper error handling and fallbacks
• Monitor performance under load
• Plan for scalability and concurrent users

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

Ready to create your first agent? Start with our Agent Creation Guide or learn more about adding content to your knowledge base.