Anzen Knowledge Base

What is the Knowledge Base?

The Knowledge Base is part of the standard Anzen platform: your team writes help articles in the management app, and anyone in the workspace can browse and read the published ones. Browsing and managing articles is included at no extra cost and gated only by the normal role-based permissions. On top of that, the optional Knowledge Base AI add-on adds a question box - "Ask your knowledge base with AI" - where a user can ask "which backup requirements apply to our CRM?" and get a short, plain-language answer grounded in your policies and the specific applications, processes and assets they are allowed to see. The article browse is free; the AI ask layer is what the add-on sells, and every AI answer cites where it came from.

How the AI ask layer works

The sections below describe the optional AI add-on. From a user's point of view the AI layer is a single question box. Behind it:

Ask. A user types a question in their own language (English or Dutch) from the Knowledge Base panel in the portal.
Gather context. Anzen retrieves the most relevant clauses from the policies that user is allowed to read, plus the applications, business processes and configuration items in their view.
Answer. Only that pre-filtered context, never your whole workspace, is sent to an EU-resident model, which drafts a grounded answer.
Cite. The answer comes back with citations: the exact policy clauses it relied on, and the applications or processes it referenced, each linking back into Anzen.

What AI answers are grounded in

The AI add-on only reasons over your own workspace data, never the open internet:

Policy clauses. Individual sections of your policies are ranked by relevance to the question, and the closest matches become the basis for the answer.
Applications. The apps in the user's view, so answers can be specific to a named system.
Business processes. The processes the user can see, so policy guidance can be tied to how work actually runs.
Configuration items. Relevant assets from the CMDB, so an answer can reference the concrete infrastructure in scope.

AI answers respect each person's permissions

The AI add-on never widens what someone can see. Context is assembled per user and permission-checked per record: a user only ever gets answers grounded in the policies, applications, processes and assets they already have read access to. Two people asking the same question can get different answers because they are allowed to see different things. A user with no access to a policy will never see its content surface in an answer.

Every AI answer is cited

Answers are not a black box. Each response lists the policy clauses it drew on, with their clause references and a link straight to the source in Anzen, plus any applications, business processes or configuration items it mentioned by name. Users can verify the answer against the underlying policy in one click, which keeps the feature trustworthy for compliance-sensitive questions.

Browsing and reading articles (included)

Knowledge Base articles are short, curated how-to pieces written by your team (for example "How to reset your password" or "Reporting a security incident"). Browsing them is part of the standard platform at no extra cost: any authenticated user can search articles by keyword, filter by category, and read the published ones. In the customer portal, end users get the same free browse over published articles. When the AI add-on is on, related articles also surface beneath each answer, matched by keyword and expandable inline. Articles are your own content and are never sent to a language model.

Writing and managing articles (included)

Authoring Knowledge Base articles is also part of the standard platform - no add-on required. Articles are written in the management app under Knowledge Base. Each article has a title, an optional URL slug (generated for you if left blank), a short excerpt shown in search results, tags, and a Markdown body. Set an article to Published to make it visible to end users, or keep it as a Draft while you work on it. Access is controlled by the Knowledge Base Article permission, granted per role in Permissions: a role can be given any of create, read, update, or delete, so you can let some teams author articles while others only read them. Every change is recorded in the audit trail.

Fair-use limits (AI add-on)

Per user. Up to 200 questions per user per day.
Per workspace. Up to 1,000 questions across the whole workspace per day.
Resets. Both counters reset at midnight UTC. The limits keep costs predictable on the included model and are comfortably above normal day-to-day use.

Pricing

Browsing and managing Knowledge Base articles is included in the standard platform at no extra cost. The optional Knowledge Base AI add-on - the "Ask your knowledge base with AI" layer with cited, plain-language answers - is €59 / month, billed in addition to your plan; it covers the AI ask experience only, not article management. Activation is self-service from the Add-ons page; cancellation takes effect immediately and stops further charges. Like every add-on it is billed on its own monthly invoice, anchored on the day of activation, and follows the same VAT rules as your plan. Workspace superadmins can also switch the AI feature off for the whole workspace from the Knowledge Base settings without cancelling the subscription.

Tips for good AI answers

Keep your policies in Anzen current. Answers are only as good as the policy clauses they draw on, so an up-to-date policy library gives sharper results.
Ask specific questions. "What are our access-review requirements for finance applications?" beats "tell me about access" and produces tighter citations.
Use it to navigate, then verify. Treat the answer as a fast pointer to the right clause, then open the cited policy for the authoritative wording.
Mind the permissions. If an answer seems thin, the user may simply not have access to the relevant policy or application yet.