“What is the composition of my monthly installment amount?” “I have a new checking account, please change my bank details.” Requests and tasks like that are addressed to Energie Steiermark’s customer service department several times on a daily basis. Usually, they are first recorded within a ticket system to ensure optimal process tracking. Depending on the topic, various employees within the company are responsible for answering them and usually several loops of processing are necessary. Hundreds of thousands of such tickets are generated each year. The knowledge contained therein will be used within the project together with Know-Center in order to optimize the customer service processing. Text preprocessing and natural language processing are the main focus, especially those in German.
Understanding the customer’s language
Finding a solution that identifies the customers’ concerns automatically and at an early stage presents a challenge. This allows inquiries to be preselected and assigned efficiently to the person in charge (skill-based routing). Dissatisfaction can be addressed proactively. In addition, it is crucial to understand the customer’s language. In each domain, there are specific technical terms with which the employees are familiar but which do not necessarily correspond to the customers’ choice of words. Repeated inquiries or difficulties in understanding are eliminated if the language styles match.
„Know-Center supports us with the continuous development of digitization in customer service. The use of NLP enables us to answer and solve our customers’ inquiries faster and more efficiently. Thanks to this advancement, we are able to respond even better and more individually to their needs and inquiries. This increases both our customer and employee satisfaction.“
Christoph Ehrenhöfer, Head of Customer Experience at Energie Steiermark Kunden GmbH, explains.
Around the clock service guaranteed by Chatbots
However, customers expect not only fast and proper service, but also around the clock availability. Chatbots that are ready for use at all times are the solution. Together with Energie Steiermark Kunden GmbH, Know-Center is exploring chatbot solutions, which can be used to quickly handle simple requests, such as providing a copy of the latest electricity bill. A database adapted that contains knowledge based on previous inquiries and replies is the chatbot’s core. Researchers use methods of similarity analysis and synonym recognition in order to identify frequently used phrases in the ticket based on the administrator’s responses. In addition, it is determined which combinations of questions occur frequently. By using prefabricated text modules, the chatbot is able to create an automated response.
As a next development step, even more complex inquiries will be answered automatically. In addition, a reference service will provide customers with helpful and continuative information or draw their attention to new offers.
The project was funded within the framework of COMET – Competence Centers for Excellent Technologies by BMK, BMDW, Land Steiermark. The COMET program is processed by FFG.