The knowledge base allows your agent to answer questions using your own data through RAG (Retrieval-Augmented Generation).

​
Source Types

File Upload

Upload PDF, DOCX, TXT, or CSV files

Website Crawl

Automatically extract content from websites

FAQ

Create structured question-answer pairs

Raw Text

Paste text content directly

​
Adding Sources

​
File Upload

1

Go to Knowledge Base

Navigate to your agent’s Playground → Knowledge Base section.
2

Click Add Source

Select File Upload from the options.
3

Upload Files

Drag and drop or select files. Supported formats:
  • PDF (up to 50MB)
  • DOCX (up to 20MB)
  • TXT (up to 10MB)
  • CSV (up to 10MB)
4

Process

Files are automatically processed and indexed.

​
Website Crawl

URL: https://example.com/docs
Max Pages: 50
Include Subpages: Yes
The crawler will:
  • Extract text content from pages
  • Follow internal links (up to max pages)
  • Skip duplicate content

​
FAQ

Create structured Q&A pairs: 
{
  "question": "What are your business hours?",
  "answer": "We're open Monday-Friday, 9am-5pm EST."
}

​
How RAG Works

  1. User asks a question
  2. Question is converted to embeddings
  3. Similar content is retrieved from knowledge base
  4. Retrieved content is added to the AI context
  5. AI generates response using the context

​
Best Practices

Add only relevant content. Too much unrelated data can confuse the agent.
Keep your knowledge base current with the latest information.
FAQs provide more precise answers for frequently asked questions.
Always test your agent after adding new sources to verify accuracy.

​
Storage Limits

PlanStorage
Free50 MB
Pro500 MB
Business5 GB