Skip to main content

Knowledge Base

The Knowledge Base is where you store information that your AI agents can reference when responding to questions or making decisions. Instead of relying only on general AI knowledge, your agents can pull from your specific documents, FAQs, and business data.

Why Use a Knowledge Base

Without a knowledge base, AI can only use its general training data. With a knowledge base:
Without Knowledge BaseWith Knowledge Base
Generic responsesResponses based on your actual products, policies, and data
Can’t answer company-specific questionsAccurately answers questions about your business
May hallucinate detailsGrounds responses in real documents
Same for every businessCustomized to your specific context

What to Store

Add information your AI agents need to do their job:
  • Product catalogs and descriptions
  • Pricing information
  • Technical specifications
  • Inventory data
  • Return and refund policies
  • Shipping information
  • Terms of service
  • Internal procedures
  • Common customer questions
  • Troubleshooting guides
  • How-to instructions
  • Known issues and workarounds
  • Company information
  • Contact details
  • Store locations and hours
  • Team member directories

Setting Up Your Knowledge Base

1

Navigate to Knowledge Base

Open your application and find the Knowledge Base section in the sidebar.
2

Add content

Upload documents or add text directly. Supported formats include:
  • PDF documents
  • Word documents (.doc, .docx)
  • Text files (.txt)
  • Spreadsheets (.csv, .xlsx)
  • Web pages (via URL)
3

Organize with collections

Group related content into collections. For example:
  • “Product Catalog”
  • “Support FAQs”
  • “Company Policies”
4

Connect to workflows

Your AI nodes (like Ask AI) automatically have access to the knowledge base. They’ll pull relevant context when generating responses.

How It Works

When an AI node runs:
  1. Query analysis — AI understands what information is needed
  2. Retrieval — Relevant chunks from your knowledge base are found
  3. Context injection — Retrieved information is added to the AI’s context
  4. Response generation — AI generates a response using both its training and your specific data
This technique is called RAG (Retrieval-Augmented Generation).
The AI doesn’t memorize your entire knowledge base. It retrieves only the most relevant pieces for each query, keeping responses focused and accurate.

Best Practices

Keep Content Up to Date

Outdated information leads to wrong answers. Review and update your knowledge base regularly:
  • Set a reminder to review quarterly
  • Update immediately when policies change
  • Remove obsolete content

Structure for Retrieval

The AI retrieves chunks of text, not entire documents. Help it find the right information:
Do: Use clear headings, short paragraphs, and organize by topic.Don’t: Upload one giant document with everything mixed together.

Be Specific

Vague content leads to vague answers.
Instead ofWrite
”Returns are handled case by case""Returns are accepted within 30 days with receipt. Items must be unused and in original packaging. Refunds are processed within 5-7 business days."
"Contact support for help""Email [email protected] or call 1-800-XXX-XXXX, Monday-Friday 9am-5pm EST”

Test Your Knowledge Base

After adding content, test with realistic questions:
  1. Ask questions a customer would ask
  2. Check if responses are accurate
  3. Identify gaps and add missing information
  4. Verify the AI cites the right sources

Managing Content

Updating Documents

When you update a document:
  • Previous version is replaced
  • Changes are immediately available to AI
  • No need to reconnect to workflows

Deleting Content

Remove content you no longer want AI to reference:
Deleted content is immediately unavailable to AI. Make sure you won’t need it before removing.

Monitoring Usage

Track how your knowledge base is being used:
  • Which documents are retrieved most often
  • Which queries don’t find good matches
  • Where you might need to add more content

Common Use Cases

Customer Support Bot

Knowledge base contents:
  • Product FAQs
  • Troubleshooting guides
  • Return policy
  • Shipping information
Result: Bot accurately answers product questions and explains policies without making things up.

Sales Assistant

Knowledge base contents:
  • Product catalog with features and pricing
  • Competitor comparisons
  • Case studies
  • Discount policies
Result: Assistant provides accurate product recommendations and can discuss specific features.

Internal Help Desk

Knowledge base contents:
  • IT procedures
  • HR policies
  • Software guides
  • Org chart
Result: Employees get accurate answers about internal processes and policies.

Next Steps