AI & RPA: End-to-end automation in call centers

A growing German bicycle retailer is struggling with the quality and efficiency of its customer service: dissatisfied customers due to long waiting and response times and simultaneously rising costs. For customer interaction, Servicetrace combined an intelligent chatbot with the RPA solution XceleratorOne. This means that many of the standard requests can be processed fully automatically – and to the complete satisfaction of the customers.

The customer service of a growing medium-sized bicycle dealer wants to optimize its call center. Due to the growth of the company and the increasing share of online sales, the number of customer inquiries is also growing. This applies in particular to questions regarding the delivery of goods. In addition, there are seasonal phases with an above-average number of enquiries. These can often not be answered on time, customers are stuck in the telephone queue. The result: poor customer ratings and negative social media feedback. The bicycle dealer was therefore urgently looking for a solution to improve call center processes. The most important goals: To increase customer satisfaction and to be able to process larger volumes of enquiries quickly without high additional costs.

The bicycle dealer’s consulting partner therefore recommended the use of an intelligent chatbot in combination with Servicetrace’s robotic process automation solution XceleratorOne (X1). Because only the coupled use of these technologies enables real “end-to-end automation”, i.e. the automation of a complete process from start to finish.

Faster processing time, higher customer satisfaction, more efficiency


For example, the bicycle dealer is now in a position to have a large proportion of standard customer service queries handled fully automatically by the chatbot. More complex inquiries are answered semi-automatically, i.e. in cooperation between the chatbot and the employee. If the chatbot is unable to assign or process a request, it forwards it directly to a call center agent.

The result of the automation with the help of the Chatbot RPA solution is impressive: Customer satisfaction increased significantly within a few weeks – more than 80% positive evaluations. The reasons for this are the faster response time (+85%), the 24/7 availability and the speedy processing. Most inquiries can be processed within a few hours. Almost every third customer request can be processed by the chatbot with the XceleratorOne solution. Further quality improvements result from the now 100% traceable customer communication, which can be analyzed more easily. These data evaluations quickly reveal optimization potentials in customer communication content and lead to continuous quality improvement.

At the same time, the call center works much more efficiently: Due to its high availability and scalability, the Chatbot solution can easily manage higher query volumes. This leads to a significant reduction in the workload of call center employees and 25% lower overall costs.

End-to-end automation with Chatbot, Artificial Intelligence & RPA

Within a few weeks, the Chatbot RPA solution was ready for use in the call center. To simplify matters, the end-to-end automation works in three steps (see diagram). The chatbot uses semantic methods such as cognitive computing and natural language processing (NLP) from the field of artificial intelligence to identify topics and categorize them. The chatbot then triggers the automated process that the X1 solution executes.

For this project, Servicetrace used the intelligent chatbot solution of its partner Cognesys, as it can be trained quickly and precisely for call center processes using cognitive methods. It is also very easy to integrate with the X1 solution via the REST API.


Example: The automated end-to-end process for express delivery requests:

  • The chatbot “reads” the input channels live chat and e-mail inbox and categorizes the requests by topic.
  • If he cannot process the request himself, he passes it on to a call center agent.
  • In the case of inquiries regarding delivery times/express delivery, the chatbot immediately asks the customer for relevant details such as the order number.
  • The X1 Bot, the software robot, then checks the request and compiles the required data with access to Salesforce and SAP
    • Salesforce: Review available product inventory and process flow
    • SAP: Check the current order status.
  • Based on the information from Salesforce and SAP, the X1 Bot is able to decide whether express delivery is possible.
  • The customer is informed about the possibility of express delivery.
  • If this is possible, the chatbot asks the customer for written confirmation.
  • The solution then creates the invoice for the express delivery, sends it by e-mail to the customer and triggers the express delivery with the responsible service provider.

Advantage of X1: highly scalable & easy Citrix integration

The customer opted for the X1 solution from Servicetrace for two technical reasons, among others:

1. because of the high and unrivalled low-cost scalability and2. the simple and effective handling of Citrix environments.

It was very important to the bicycle dealer to find a solution that could also manage high volumes of enquiries without additional effort. XceleratorOne offers a unique scaling approach with a multiple patented technology: vertical scaling. This means that X1 can provide significantly more processing capacity on the existing infrastructure. This is an extremely cost-effective approach compared to the competition (more information on this scaling methodology can be found here).

In addition, the bicycle dealer used the Citrix virtualization platform for its IT infrastructure, which can be a challenge for robot-controlled process automation. But not for X1: The simple work with Citrix is one of the main distinguishing features of the service trace solution. Using pattern-based recognition and integrated Optical Character Recognition (OCR), X1 is able to control Citrix desktops and applications as effectively as any other application, regardless of the end device.


Read more case studies and customer references about Robotics Process Automation here.