Process escapades | Part 2: Service desk
Process escapades that you (don’t) want to experience
Inspired by the well-known book series “30 Escapades You Must Experience”, we take you on a special journey through everyday working life—to places where processes fail, chaos reigns and employees struggle with problems that could have been solved long ago. These are not adventures that you seek out voluntarily. But they are precisely the challenges that many companies have to deal with every day.
We show you three typical escapades that are exemplary for many: sometimes funny, sometimes chaotic, but always with real insight value. And we’ll show you how structured, digital and partially automated BPMN processes can be used to create clear, comprehensible and efficient workflows even from muddled situations—with tangible added value for everyone involved. Because if you really understand processes, you can not only improve them, but also use them in a targeted manner to relieve employees, ensure compliance and increase satisfaction in day-to-day work.
Our stops on this journey:
- From the authorization chaos in HR service, which can only be tamed with structure and clarity.
- From ticket routing in the service desk, which finds a new way between stray tickets and AI intelligence.
- Right through to the ordering process, where an absurd helmet requirement suddenly becomes a serious process blockage.
Part 2: Ticket routing between ticket strays and intelligence
A field report from the service desk
In the hustle and bustle of a modern IT service desk, a small click often decides whether a process is resolved quickly or delayed for days. What sounds banal at first—the correct classification of a ticket—has far-reaching consequences in reality.
Time and again, requests end up in the wrong department. Tickets move back and forth between different processors like a game of ping-pong. Every forwarding means a loss of time, queries, duplicate processing—and in the end: frustration for everyone involved. End users wonder why simple issues are not resolved quickly. IT staff lose time re-sorting and redistributing. And for the company, the result is a process that is neither efficient nor evaluable.
The reason for this chaos usually lies in a manual step: categorization. In many organizations, this is done by hand at the start of a workflow—under time pressure, with incomplete information, by employees who are not always sure which label is correct. Errors are inevitable. And that has consequences: Not only does the average processing time increase, but the quality of the process data also suffers. Incorrectly classified tickets distort any subsequent analysis. Improvement initiatives are then based on an unclean data foundation.
This systematic lack of clarity paralyzes further development. Because as long as it is not clear where bottlenecks really occur or which issues are frequently assigned incorrectly, any optimization remains in the fog. At the same time, satisfaction within the team decreases—both among IT employees, who struggle with routine issues, and among end users, whose concerns remain unresolved for too long.
The turning point: AI meets BPMN
To break out of this looping process, we have completely rethought the workflow. The approach: not an isolated solution, but the combination of intelligent automation with a structured process model.
At the heart of this is an AI-based classification system that has been seamlessly integrated into the existing ticket system workflow. As soon as a request is received, the AI analyzes its content, recognizes whether it is a incident or a service request, checks whether there is a suitable standard solution—and suggests the appropriate ticket category. This decision is not arbitrary, but is based on trained data, empirical values and defined probabilities.
The BPMN process picks up on these predictions, transfers them into defined workflows and automatically controls the routing to the right processing teams. Non-technical requests—such as tickets that actually concern the facility service - are automatically recognized and forwarded without the need for human intervention. This not only eliminates the biggest uncertainty factor, but also a considerable amount of manual effort.
Acceptance within the team was surprisingly high. The employees did not experience the AI as a controlling authority, but as a helpful relief. They save themselves monotonous assignment work and can concentrate on the relevant cases. End users, in turn, benefit from faster responses and more precise routing.
One IT employee put it in a nutshell: “The AI takes over the time-consuming tasks—and we use the time gained for more meaningful topics.”
Result: clear course, measurable impact
Since the introduction of the new process, the average throughput time for tickets has dropped noticeably. Requests reach the right contact person more quickly, categorization errors have been significantly reduced and data quality has improved massively.
The combination of AI-based intelligence and a clear BPMN structure ensures a system that is capable of learning—while remaining controllable. The results can be checked, processes can be adapted and the entire organization benefits from more transparent processes.
Conclusion on the second escapade
A process does not have to be complex to be effective. But it does need to be set up intelligently. The combination of structured BPMN modeling and automated AI support has turned an inefficient and error-prone help desk process into a modern, powerful solution.
The errors have disappeared—and the satisfaction has reached its goal.
In part 3 of our series, things get more tangible: When a helmet requirement in the ordering process suddenly becomes a process blocker, it shows how important practicable exceptions and clean BPMN processes can be in procurement.
Part 3 follows: Helmet obligation in the ordering process
.