AI agent knowledge

In this article, you'll learn how knowledge sources work in Enjo, how to add and manage them for your AI Agent, and which integrations are supported.

How knowledge works

Each AI Agent is grounded in one or more knowledge sources. When a user asks a question, the AI Agent searches connected knowledge sources, generates a response based on the most relevant content, and includes citations with links back to the source.

Knowledge sources can contain pages, documents, articles, or tickets from a connected app. A single knowledge source can span multiple folders, spaces, or categories within that app.

Global vs agent-specific knowledge

Global knowledge sources are managed centrally and can be shared across multiple AI Agents. Configure these from the Global Knowledge page in your workspace settings. This is useful when multiple agents need access to the same content — for example, a company-wide policy document.

Agent-specific knowledge sources are configured directly on an individual AI Agent and are only available to that agent. Use these when an agent needs access to specialized or team-specific content.

💡 Start with global knowledge sources for content that applies across your organisation. Use agent-specific sources to layer in content that is only relevant to a particular team or use case.

Add a knowledge source

1. Go to AI Agents and select the agent you want to configure.

2. Open the Knowledge tab.

3. Click Add Knowledge.

4. Select the app or source type from the drawer.

5. Follow the setup steps for your selected source. See the integration guides below.

Preferred knowledge sources

You can set a preference order for knowledge sources on an AI Agent. When multiple sources contain relevant content, the AI Agent prioritizes sources higher in the preference order.

Supported knowledge sources

💡 The integration surface expands continuously. If a source you need is not listed, contact the Enjo support team.

What's next

• Guidelines for Knowledge— Manage shared knowledge sources across your workspace.

• Confluence as Knowledge — Ground your AI Agent in your confluence

• Notion as Knowledge — Validate AI responses across your knowledge sources at scale.