Skip to main content

Knowledge Base

The knowledge base lets you upload documents that agents can search and reference during conversations. When a caller asks a question, the agent retrieves relevant passages and uses them to construct an accurate answer.

How it works​

  1. Upload documents (PDF, DOCX, TXT) to a project's knowledge base
  2. Vois AI chunks and embeds the content using vector search
  3. When an agent needs to answer a question, it queries the knowledge base and retrieves the most relevant passages
  4. The agent uses those passages as context when generating a response

This is called Retrieval Augmented Generation (RAG) — the agent is grounded in your documents rather than relying solely on the LLM's training data.

Adding documents​

  1. Go to your project → Knowledge Base
  2. Click Upload Document
  3. Select a PDF, DOCX, or TXT file (max 50 MB)
  4. Wait for processing — large documents may take a minute

Enabling knowledge base for an agent​

Open the agent editor → Tools tab → enable Document Lookup.

The agent will automatically search the knowledge base when it needs information it doesn't have in its system prompt.

Best practices​

  • Use clear headings and structure in your documents — chunks with good context retrieve better
  • Upload focused documents rather than large omnibus files
  • Keep documents up to date — outdated content can mislead the agent
  • Test retrieval in the playground after uploading new documents