What Are Knowledge Sources and Why Do They Matter?
Your agent is only as good as what you give it to learn from.
Table of Contents
You have already done most of the work. The documentation exists. The articles are written. The answers are somewhere in your systems. The problem is that your customers cannot find them, and your team ends up answering the same questions manually anyway.
It is Monday morning. A customer asks your agent how to export their data. Your agent finds the answer instantly - pulling it from the exact article in your knowledge base that covers it - and delivers a clear, accurate response in seconds. The customer gets what they needed without waiting, without a ticket, without anyone on your team having to get involved.
That is what connected sources make possible. Knowledge Sources link your existing content to your agent so it can draw from everything your team has already documented. Connect the right sources and your agent becomes a reflection of everything your organisation knows - available to every customer, at any hour, automatically.
In this article, you'll learn:
- What knowledge sources are
- The different types of sources available
- How your agent uses sources to answer questions
- How to get to the Sources tab
What Is a Knowledge Source?
A knowledge source is any piece of content you connect to your agent. When a user asks a question, your agent searches through all connected sources to find the most relevant answer and responds based on what it finds.
Sources can be:
- Your Helpjuice knowledge base
- Third-party tools like Confluence, Google Drive, or Intercom
- Files you upload manually (PDFs, Word docs, spreadsheets, and more)
- Websites you want the agent to crawl
- Short written answers you create directly in Outlearn, called Snippets
Types of Sources
| Type | What it is | Best for |
|---|---|---|
| Snippets | Short answers you write directly in Outlearn | FAQs, quick facts, canned responses |
| Knowledge Base | Your Helpjuice KB or other help doc platforms | Existing help content and product documentation |
| Cloud Storage | Google Drive, Dropbox, Box, OneDrive | Internal documents, SOPs, policies |
| Ticketing Systems | Freshworks, Intercom, Zendesk | Past support conversations as context |
| Team Chats | Slack, Teams, Google Chat, Gmail | Internal knowledge and team communications |
| Files | PDFs, DOCX, CSV, XLSX, TXT, MD, HTML | One-off documents without an integration |
| Website URL | Any publicly accessible webpage | Product pages, marketing content, public docs |
How Your Agent Uses Sources
When a user sends a message, your agent doesn't just search for keywords - it understands the meaning of the question and finds the most relevant content across all connected sources.
The process looks like this:
- User asks a question.
- Agent searches all active sources simultaneously.
- Agent finds the most relevant content.
- Agent writes a response based on that content.
This means the quality and relevance of your sources directly affects the quality of your agent's answers. Well-organized, up-to-date content produces accurate answers. Outdated or thin content produces weak ones.
The Sources Tab
All source management happens in the Sources tab in the top navigation bar.

From here you can:
- Add new sources and snippets
- Toggle sources on or off
- Manage accessibility settings
- Resync or delete sources
Best Practices
- Connect your most important content first - your top 10–20 most-viewed knowledge base articles are a better starting point than connecting everything at once.
- Quality beats quantity. A focused set of accurate, well-written sources produces better answers than a large set of outdated or irrelevant content.
- Use the Content Gaps section in Analytics to see what questions your agent couldn't answer - it tells you exactly what sources to add next.