Webinar overview | Break down knowledge silos in education with XWiki and Collabora Online

XWiki does not have a centralized moderation system that automatically applies the same anti-abuse rules to every instance. Each XWiki deployment can have different requirements depending on its use case, infrastructure, and expected workload.

For a self-hosted XWiki instance, policies, protections, resource limits, and other safeguards can be configured together with your infrastructure team to match your specific needs.

For XWiki Cloud, if you expect unusually large volumes of content, traffic, users, or automated activity, it is best to discuss the use case with your Account Manager in advance. The team can then determine which configuration and limits are appropriate for the deployment.

This means legitimate high-volume use cases do not have to fit into a single, platform-wide moderation model.

For a sovereign and cost-conscious setup, XWiki can be self-hosted, giving your organization control over its data, infrastructure, and deployment.

For an internal job board, a practical approach is to use App Within Minutes (AWM) to create a dedicated job application. You could define fields such as job title, department, location, status, or description. Each vacancy will be stored as a separate XWiki page and all positions listed together on a dedicated application home page.

You can also use XWiki permissions to control who can access individual job pages. For example, some positions could remain visible only to employees, while others could be made available to guest users.

Collabora Online can complement this setup when teams need to work collaboratively on office documents associated with those processes, while keeping those documents connected to the relevant XWiki pages.

XWiki is well suited to this type of use case because information does not have to be stored only as plain text. Pages can also contain structured data and metadata.

At a general level, you can associate an XWiki Object with a page and use it to store specific properties, such as an EIN, ABN, ISBN, ISAC, or other identifiers. This makes the metadata part of the content model rather than simply text written somewhere on the page.

A practical way to build this is with App Within Minutes (AWM), which lets you create custom entity types with defined properties. Those entities can then be created, linked, and reused throughout the XWiki instance, making it possible to build a more structured university or organizational knowledge hub.

For OpenProject, yes. XWiki provides a dedicated OpenProject Integration (Pro) that connects an XWiki instance to OpenProject and lets users work with OpenProject work packages from XWiki.

For Zammad, there is currently no equivalent XWiki SAS connector documented in the XWiki extension catalog. Zammad does, however, provide a REST API covering tickets, users, organizations, its knowledge base, and other objects. Therefore, a custom integration with XWiki can be developed depending on the workflow required.

XWiki provides a dedicated Confluence Migration Toolkit designed to migrate content together with information such as page history, users, permissions, attachments, and metadata. It also includes tools and compatibility extensions intended to reduce the amount of manual rebuilding after migration.

The current trial of the Confluence Migrator allows up to 200 pages per migration, so it can also be used to evaluate a representative part of an existing Confluence instance before planning the complete move.

XWiki has now worked with more than 100 organizations on Confluence-to-XWiki migrations, which has also shaped the migration tooling and methodology used today.

For version control, every change to an XWiki page is recorded in its history. Users can inspect earlier versions, compare changes, and roll back when necessary. XWiki also keeps versions of attachments by default, which is useful when office documents are stored as page attachments.

For searching across the wiki, XWiki indexes page content and provides full-text search and filtering. If the intention is simply to find a word inside the page currently open in the browser, the normal browser Find command, Ctrl+F on Windows/Linux or Cmd+F on macOS, is the simplest option.

For editable structured fields, App Within Minutes lets you define fields such as short text, long text, numbers, dates, users, lists, and other properties. Application entries can then be edited by end users who have the appropriate permissions, including in-place editing for supported fields.

With Collabora Online, you can also enable version control within documents. This allows you to track and manage changes for all or specific users, as well as compare documents.

Yes. For dynamic data, the recommended approach is Live Data rather than a static wiki table. Live Data supports sorting by multiple properties, filtering, pagination, different layouts, and, where supported, inline editing. Users can change the sort order directly from the displayed data.

The demo used XWiki's diagram functionality. The supported Diagram Application (Pro) integrates diagrams.net, also known as draw.io, into XWiki. Diagrams can be created and edited inside XWiki, included in other pages, versioned, and imported from formats such as draw.io and Gliffy.

Yes. When Collabora is used through XWiki, the document follows the permissions of the XWiki page to which it is attached. A user needs the corresponding XWiki permissions in order to view or edit the attached office document.

Yes. XWiki Business Applications, including the Collabora Connector Application (Pro), can be tested with a free trial license. The connector's Store page provides the installation and trial-license process.

For Collabora, CODE is free and can be installed locally using Docker, Linux packages, a virtual appliance, or other supported methods. It is intended for testing, development, home use, and small teams rather than production environments.

For AI, Collabora Online is provider-agnostic. The AI assistant works with any LLM provider that exposes an OpenAI-compatible endpoint, whether that's a commercial provider or a self-hosted model such as Ollama, vLLM, and LM Studio. Models used for document actions need native tool/function-calling support, while models without it can still be used for simpler chat interactions.

Collabora Online supports all major document, spreadsheet, and presentation formats including Microsoft formats such as DOCX, DOC, XLSX, XLS, PPTX, and PPT, as well as OpenDocument formats. It also provides import or viewing support for formats such as PDF, Markdown, Visio, and Publisher.

There is no need for an immediate conversion project. Existing Microsoft Office files can continue to be edited and created in Collabora Online. During the webinar, Richard specifically recommended open formats while emphasizing that existing DOCX and similar files can continue to be used.

Collabora Online is open source, can be deployed on premises, and integrates into other platforms rather than requiring organizations to move their files into a separate proprietary storage environment. It provides a feature-rich, browser-based, collaborative editor for documents, spreadsheets, presentations, and drawings, supporting both ODF and widely used Microsoft Office formats. There is also a dedicated page where you can learn more about how Collabora compares to other collaboration suites.

For organizations already using XWiki, the important advantage is the integration itself: office documents can remain attached to the knowledge they belong to instead of becoming a separate content silo.

Yes. Writer includes Track Changes, with the ability to record edits, identify their authors, review them, and accept or reject individual changes.

To begin tracking changes in your document, navigate to the "Review" tab and click "Record". Automatic spell checking is also available in Writer, Calc, and Impress.

The model used during the demonstration is DeepSeek. More importantly, the Collabora AI assistant is not tied to that model: administrators can configure their own provider or self-hosted endpoint.

Collabora itself does not collect or retain the document through the AI assistant. The assistant has no document access by default, and document content is shared only when the user explicitly permits it.

Once information is sent to an external AI provider, however, retention depends on the provider and configuration you selected. If this is a concern, organizations can connect Collabora to a self-hosted model so document data remains within infrastructure they control. This distinction is important: Collabora controls when data is shared, but the retention policy after that point is determined by the chosen AI endpoint.

Yes. XWiki can be deployed on premises, Collabora Online can be deployed on premises, and the Collabora AI assistant can connect to a locally hosted LLM. This allows the complete workflow to remain inside infrastructure controlled by the organization if required.