On 27.11.2018 OMQ successfully held the first workshop of the series “Customer Centricity” in Berlin.
The participants there discussed central topics of customer service and the use of artificial intelligence. The audience was made up of various sectors such as e-commerce, software, publishing, insurance and public institutions.
Guest speaker Manuel Kester, COO of simplesurance GmbH, explained in his field report on the use of OMQ how more than 70% of service requests in customer service are intercepted by simplesurance. The highlight of the official part was the live integration using the example of the contact page of DeinHandy.de. In this “Live Hack” the participants could see how OMQ Contact is integrated into a contact form within a few minutes and works immediately in real time.
As a prelude to the workshop, Matthias Meisdrock, Managing Director of OMQ GmbH, gave a technical introduction to the topic.
The introduction also focused on the challenges facing today’s enterprise customer service departments. The biggest problems are often poor service due to incorrect information, different answers and long waiting times. The one who suffers most is always the end customer. But also the service employees suffer from bad working conditions, because the work is characterized by high manual effort. Due to the high fluctuation rate, companies incur high training costs. In addition, different systems are often used, which means that many different isolated solutions are used in the individual departments.
But what does all this have to do with customer centricity?
"Customer Centricity means taking care of the customer and placing the customer at the center of your products, also in service," says Matthias Meisdrock.
For OMQ this means that the customer always gets a suitable, high-quality answer and that he also receives it on every channel in real time. The OMQ vision is to become the center of the AI-based automated service. OMQ functions as a holistic system. The focus is on the intelligent knowledge database, which distributes the knowledge to all channels and answers queries. As an answering contact form, dynamic FAQ, self-answering e-mail system, helping chat bot and supporting ticket system, up to 80% of customer inquiries are intercepted.
This is achieved by a technology developed by OMQ. This differs from other technologies in three special characteristics.
1. Self-learning technology
The first feature is that OMQ is a self-learning system. It works without introduction and training. In the process, texts from customers are automatically viewed and it examines which relationships certain words have to each other. For this purpose, word concepts are created that show how the customers express themselves. In addition, by tracking the click behavior of employees and customers, the system analyzes which answer helped the customer. So the system gets better with every request.
2. Hybrid Learning Supervised/Unsupervised
The second feature of the technology is that it is a hybrid system that learns both unsupervised and supervised at the same time. Unsupervised means that it learns from blank user behavior data as described above. Superviesd, on the other hand, means that it is additionally trained (independently with OMQ). New word concepts emerge again and again, which change continuously due to the influence of customer feedback. Thus the system can train itself again and again, and assign words to other concepts. Based on the word concepts, question concepts are then formed, which are also constantly improving.
3. Questions learn from each other
The third characteristic is that questions learn from each other. A big advantage of the OMQ procedure is that new questions can be easily added to the database. For this no new training is necessary. Existing concepts are used. The system is thus able to form associations automatically and can be expanded continuously and dynamically. This way more and more questions can be answered.
So much for the first part of the OMQ Workshop “Customer Centricity”. The video of the field report and the live integration will follow shortly.