Knowledge Base / RAG

Give your agents context about your business by uploading documents. AgentJara uses Retrieval-Augmented Generation (RAG) to find relevant information.

Supported Formats

  • Documents — PDF, DOCX, TXT, Markdown
  • Web pages — Paste any URL and we'll extract the content
  • Spreadsheets — CSV, XLSX
  • Raw text — Paste text directly

How RAG Works

When you upload a document, AgentJara splits it into chunks, generates embeddings, and stores them in a vector database. When your agent receives a question, it searches the knowledge base for relevant chunks and includes them in the context.

Best Practices

  • Upload your FAQ documents first — they have the highest impact
  • Keep documents focused and well-structured
  • Update knowledge regularly as your business changes
  • Use the knowledge playground to test retrieval quality
← AgentsNext: Models →