I’d like to hear from the TOPdesk community about your experiences with generating knowledge items using AI.
In our experience, the AI often makes the resulting content too short. Even when we provide a detailed solution summary with multiple steps, it tends to generate either:
- A very brief summary or short explanation; or
- A numbered 1–2–3–4–5 step plan containing only headlines, without enough detail or context.
This means we often need to expand and rewrite the generated knowledge items manually.
Have you experienced the same issue? If so, have you found any effective prompts, settings, instructions, or other tricks to make the AI generate more complete and detailed knowledge items?
I’d be particularly interested in tips for ensuring that the AI:
- Includes all relevant solution steps;
- Provides sufficient explanation for each step;
- Maintains the desired level of detail;
- Produces a consistent knowledge item structure; and
- Avoids oversimplifying the source information.
Any experiences or recommendations would be greatly appreciated.