Practical example and current challenges
- Organisations providing helpdesk services are increasingly facing requests from users due to the increase in digital services. This leads to a high number of tickets generated. Solutions to deal with this high number of tickets often result in a multilevel helpdesk approach. The classical approach is that recurring questions (whose solution is already known) are physically answered by so-called "agents" in level 1. The involvement of so-called subject matter experts is not necessary. They are usually only used in level 2 (for detailed questions that the agent cannot answer, but no specialised expert knowledge is required) or level 3 (for questions that require specialised expert knowledge to answer) - specifically, when it comes to solving more complex tasks.
- Although this organisational approach helps to manage high ticket numbers, it is still necessary to increasingly integrate agents as an additional personnel "resource" for the helpdesk due to the large ticket numbers.
- The use of "intelligent" chatbots across all levels already allows the majority of requests to be solved and answered in a dialogue-based manner. However, the quality and performance of an intelligent chatbot is directly related to the training history and the content of the content management system. In order to be able to react to questions with adequate answers and solutions, the chatbot must be trained in advance, filled with knowledge and ideally prepared for potential formulations and questions.
- By regularly analysing the helpdesk tickets - and the resulting solutions - it is possible to enrich the chatbot with adequate knowledge and ensure a qualitative dialogue-based answer to the question by the chatbot. By continuously training the chatbot on the basis of regular ticket analysis, it is also possible to drastically reduce the number of tickets. In this way, the higher quantity requirements for agent workstations can be addressed and the personnel capacities for the agents can be reallocated towards subject matter experts.
What added value does the "GAIA-X project" offer?
- By using the Gaia-X infrastructure, it is possible to share trained content across organisational boundaries. This enables an enormous increase in the quality of the chatbot as well as a better differentiation from competitors.
- Through the networking within the framework of the Gaia-X infrastructure, it is possible to train the "AI" of the chatbot across companies. This means that redundant training processes, which each organisation would have to carry out separately, can be avoided.
- Gaia-X allows the services for the chatbot to be provided both on premise and cloud-based in a data centre within the EU.
- Gaia-X's GDPR-compliant framework enables the public sector to use chatbot services in a le-gally compliant manner and to benefit from qualitative AI chatbot logic.
Use Case Team
- Christian Schieb – Unisys Austria GmbH
- Robert Kamrau – Unisys Germany GmbH