With the Confluence integration, you can bring knowledge from your company’s Confluence spaces directly into NEXT chats. This allows your team to enrich insights from interviews, calls, or surveys with background knowledge like research reports, stakeholder information, or process documentation stored in Confluence.
Confluence is often the “long-term memory” of a company. While NEXT focuses on highlights, insights, and themes, sometimes the full picture requires context from outside sources. This integration makes it seamless to pull in exactly the Confluence content you need — at the moment you need it — without leaving NEXT.
How to connect Confluence
Open the Settings dialog in NEXT.
Go to the Integrations tab.
Create a new subscription of type Confluence MCP.
Once saved, the connection is active for everyone in your teamspace.
That’s it. From now on, whenever you create a new chat thread, NEXT knows it can look up additional knowledge in Confluence when relevant.
How it works behind the scenes
When you start a thread, NEXT may realize it needs more context. This can happen in two ways:
You explicitly ask for Confluence data (“Search Confluence for …”).
Or the AI detects a gap and decides to check Confluence automatically.
Here’s what happens step by step:
Search request: The chat sends a search query to Confluence. The query is derived from what you typed. You can see the actual search terms used in the chat UI.
Result list: Confluence returns matching pages with titles and snippets. At this point, the chat doesn’t yet know the full content.
Page selection: The chat picks the most relevant page(s).
Content load: It fetches the full page content from Confluence and brings it into the chat’s context.
Answer generation: The chat uses that content, alongside your other data in NEXT, to craft a response.
If more than one page is relevant, it can repeat this process automatically.
👉 In the Thread UI, you can see a transparent log of these steps: the search request, which pages were found, and which ones were opened.
FAQ
Q: Why don’t we use Atlassian’s out-of-the-box MCP server?
Atlassian offers its own MCP server for Confluence, but it’s designed for individual workflows. In that setup, whenever the chat needs data, you (the user) are prompted in the chat to connect your personal Confluence account via a login flow. Only after that step can the chat fetch data.
This doesn’t fit how NEXT works:
Hidden knowledge surfaced: For example, a product manager might not know that research reports exist in Confluence. With teamspace-level setup, the research team can make that knowledge available without extra steps.
No Atlassian account required: Some NEXT users may not have a Confluence account (or don't have access to this part of Confluence) but still need access to the content.
Automated Threads: NEXT can also create threads automatically via Automations (without a person manually starting them). In those cases, there is no human in the loop to enter login credentials - so a user-centric Atlassian login flow wouldn’t work.
This approach ensures smoother collaboration: teams set up the connection once, and everyone in the teamspace can benefit - instantly and transparently.