Skip to main content

Fix the basis: how to build a knowledge system your team — and your AI — can rely on

  • July 28, 2026
  • 0 replies
  • 19 views

Sanne van Opstal-Brakel
Employee
Forum|alt.badge.img+15

Knowledge management isn't a one-off project, it's a habit. At most service desks, the knowledge base just kind of happens: someone writes an answer, a colleague writes a slightly different one, and before you know it ten people are working from ten versions of the same answer. That costs time — operators rewrite answers that already exist — and it costs consistency, since the same question gets a different answer depending on who picks it up. New colleagues also take weeks longer to become productive. And now that AI reads along and helps decide more and more, that problem only gets bigger: a messy knowledge base doesn't just slow your team down, it makes AI useless.

Getting the basics right starts with one simple question: what do you do when a new ticket comes in? The answer should always be the same. First, search the knowledge system — not as a second step, but as the starting point. Found it? Use the knowledge item and link it to the ticket. Didn't find it? Solve the ticket, and add a new knowledge item while you're at it, so the next colleague has the answer ready to go. That last part — adding knowledge the moment you solve something new — is where most teams get stuck. It only becomes a habit once it's built into the daily workflow, not a separate task on the side.

Structure helps a lot here. In the new Knowledge Base explorer, you set up folders and knowledge items exactly the way your team works: by department — IT, Facility Management, HR, Application Management — or by service and category, such as Accounts & Access with Password Reset and Software & business applications underneath. Which approach fits best depends on how your team already works. What matters more than the exact breakdown is that you choose deliberately, instead of sticking with a structure that just grew that way over time.

To get that foundation right, we use BPKB — Best Practice Knowledge Management — at TOPdesk. It rests on three pillars: language, meaning how you write and title knowledge items so they match how people actually search and think; people, meaning who's responsible for what and how you organise ownership around that; and process, meaning how knowledge becomes a fixed part of the daily workflow instead of something extra on top.

And that foundation only becomes more important. This summer we launched three new AI features that build on it. An AI answer assistant suggests a ready-to-use answer based on your existing knowledge items, added with a single click — the better your knowledge items, the better those suggestions get. An AI writing assistant rewrites any draft to make it clearer, shorter, or more professional, for both knowledge items and ticket replies. And AI translation lets you read any ticket in your own language with one click, no matter what language a colleague or customer originally wrote it in. That builds on what we already launched earlier this year. AI knowledge item creation arrived in March: it automatically suggests a new knowledge item based on a resolved ticket — that exact step where most teams get stuck, now made a lot easier. AI incident categorization, which automatically suggests the right category for a new ticket, and AI summary, which generates a clear overview of a ticket or conversation in one click, have been around even longer.

The better your knowledge system, the better these features work. Not sure your foundation is ready for that yet? With the AI Readiness service, now available via my.topdesk.com, we help you get your knowledge system AI-ready, with your Account Manager tailoring the package to your situation.

But whether you use that service or not: that foundation — structure, language, people, process — is something you build yourself, and you can start on it today. Five things to get started with:

  1. Agree on a clear folder structure — or take a critical look at whether your current one still holds up.
  2. Map out your 10 most common tickets and create knowledge items for them.
  3. Turn "search the knowledge system first" into a fixed habit in your team's daily workflow.
  4. Explore the new Knowledge Base explorer in your own TOPdesk environment.
  5. Need help? Ask your Customer Success Manager about the AI Readiness service.

Want to see all of this in action, including a live demo of the Knowledge Base explorer? Watch the full webinar here.

How is your knowledge base set up right now, and what's costing your team the most time these days? Feel free to share below.