Knowledge Base

The Knowledge Base feature allows you to create, manage, and utilize business knowledge and documentation that your AI agents can access to provide accurate, consistent responses to user queries.

Overview

Knowledge bases are collections of documents and information that:

• Provide factual grounding for your AI agents
• Enable accurate responses to customer questions
• Reduce the need for human intervention
• Ensure consistency in information delivery
• Scale your support capabilities efficiently

Creating and Managing Knowledge Bases

Knowledge Base Dashboard

The main dashboard shows all your available knowledge bases:

• View existing knowledge bases
• Monitor document count and total size
• See when knowledge bases were last updated
• Create new knowledge bases with the "New Knowledge Base" button

Knowledge Base Details

Each knowledge base contains crucial information:

• Name and description of the knowledge base
• Number of documents and total size
• Creation and last update dates
• Options to export or upload new documents

Document Management

Adding Documents

You can add documents to your knowledge base in several ways:

• Upload documents directly through the interface
• Documents are automatically processed and indexed

Supported File Types

The knowledge base system supports the following document formats:

• PDF Documents (.pdf)
• Microsoft Word Documents (.doc, .docx)
• HTML Files (.html, .htm)
• Markdown Files (.md, .markdown)
• Plain Text Files (.txt)

Each file type is automatically processed using the appropriate extraction method to ensure optimal content retrieval.

Processing and Indexing

When you upload a document:

1. The system automatically extracts and processes the text based on the file type
2. Content is indexed using embeddings for semantic search
3. The document becomes searchable by your AI agents
4. Documents are typically available for querying within minutes

Document Context Menu

Each document in your knowledge base has a context menu with useful options:

The context menu provides these actions:

• Test Query: Run a test query against this specific document to verify information retrieval
• Delete: Remove the document from your knowledge base

These options make it easy to manage individual documents and test how well your knowledge base responds to queries.

Using Knowledge Bases with Agents

Knowledge bases become truly powerful when connected to Agent nodes:

When configuring an Agent node:

1. Enable the Knowledge Base tool
2. Select which knowledge bases the agent should access
3. The agent will automatically search these knowledge bases when answering user questions
4. Results are ranked by relevance and used to formulate accurate responses

Common Use Cases

• Customer Support: Provide accurate answers to product questions, troubleshooting guides, and policy information
• Internal Documentation: Help employees find information in company manuals, procedures, and guidelines
• Product Information: Share detailed specifications, compatibility information, and usage instructions
• Training Materials: Make training content easily accessible through conversational interfaces
• FAQ Automation: Automatically answer frequently asked questions using your documented responses

AI Support Agent Implementation

The Knowledge Base feature is a core component of the AI Support Agent solution. By combining Knowledge Bases with Agent nodes, you can create a powerful 24/7 customer support system that:

• Automatically answers customer questions using your documentation
• Provides accurate, consistent responses based on your knowledge base
• Scales your support capabilities without increasing staff
• Handles customer inquiries even when your team is offline

To implement an AI Support Agent:

1. Create a comprehensive knowledge base with your support documentation
2. Set up an Agent node with the Knowledge Base tool enabled
3. Configure the agent with an appropriate customer support template
4. Deploy your agent to start handling customer inquiries

For a complete implementation guide, see the AI Support Agent documentation.

Best Practices

1. Organize content logically: Create separate knowledge bases for different topics or products
2. Keep information current: Regularly update documents to ensure accuracy
3. Use clear, concise language: Well-written documents lead to better AI responses
4. Include common questions: Add documents that address frequently asked questions
5. Test your knowledge base: Ask various questions to ensure the system retrieves relevant information
6. Start with high-priority content: Focus on adding the most frequently needed information first
7. Review and refine: Analyze which queries receive poor responses and improve your documentation

Example Setup

Product Support Knowledge Base

Name: Wedding Story Support
Description: Everything you need to know about our business
Documents:
- Product Catalog.pdf
- Troubleshooting Guide.pdf
- Pricing Information.pdf
- FAQ Document.pdf

Access: Connected to customer support agent

Internal Procedures Knowledge Base

Name: Internal Procedures
Description: Company policies and procedures for staff reference
Documents:
- Employee Handbook.pdf
- Technical Documentation.pdf
- Process Workflows.pdf

Access: Connected to internal team assistant

Notes and Limitations

• Knowledge base effectiveness depends on the quality and comprehensiveness of your documents
• Very technical or specialized content may require more precise phrasing in user questions
• The system works best with well-structured, clear documentation
• Large documents are automatically chunked for processing
• Knowledge bases have storage limits based on your plan
• Languages other than English may require documents in that language for optimal performance