Answers Integration - Confluence


Moveworks Answers skill enables the Moveworks bot to answer user’s questions directly in chat, by understanding the questions and responding instantly with answers containing the most relevant article snippets, links to articles in the knowledge base, or links to external articles taken from trusted knowledge sources. The Moveworks Confluence integration is leveraged by the Moveworks Answers skill to import content into the Moveworks Enterprise Cache so it can be searched by the Moveworks bot in real-time.

Supported Confluence Versions

Moveworks supports the following versions of Confluence:

  • Confluence Cloud
  • Confluence On-Prem
    • Confluence Server
    • Confluence Datacenter

Note: Confluence Server requires leveraging the Moveworks Agent to securely connect to your Confluence On-Prem.

Supported Content Types

Moveworks supports ingesting Confluence wiki articles.

Knowledge Ingestion Filtering Methods

Moveworks can be configured to filter the knowledge being ingested by Spaces or by a CQL query, typically from an ancestor page. Depending on what version of Confluence Server is supported in your system, the method of filtering knowledge ingestion will differ.

Method 1: Ingest by Spaces

In the majority of cases, Moveworks will import articles from Confluence by Space. For example, if you have a space called “IT Self Service”, Moveworks would ingest all published pages under that space.

To ingest articles by Space, Moveworks uses an API call leveraging the content endpoint like so: /wiki/rest/api/content?limit=25&status=current&spaceKey=IT%20Self%20Service .

Method 2: Ingest using Confluence Query Language (CQL)

In some cases, a given space may contain some pages should not be ingested. In this case, Moveworks can leverage the Confluence Query Language (CQL), which uses a SQL type syntax, for advanced search and filtering. One of the best parts about CQL is that it allows you to define more advanced search options such as cherry picking the content id of “ancestor” pages and recursively ingesting all the child KBs under that ancestor page.

To do this, Moveworks would use an API call that leverages the “search” endpoint to pass in a CQL query like so: rest/api/content/search?cql=ancestor=1234567859&limit=300&label="mw-inc”

Another popular method of ingestion that leverages CQL is to ingest articles by labels. In Confluence, labels are key words that you can add to pages and attachments to make them easier to group and find. Labels are similar to tags, you can label an article moveworks and then in your Moveworks knowledge ingestion configuration you could use a CQL query like: label = moveworks AND space = IT

The screenshot above shows an article with the labels “accounting” and “paid”.