ADVANCING TOTAL EXPERIENCE THROUGH ARTIFICIAL INTELLIGENCE: AN OVERVIEW

Do you remember who the Luddites were?

They were the "machine breakers" who fought against machine production in the early 19th century. They feared for their jobs, and—superficially considered—not without reason.

We know how the story ends. Many factories are now almost fully automated and still require well-trained personnel. These workers no longer stand as frequently on assembly lines but instead develop, plan, maintain, and optimize the machinery.

The transformation in communication, coupled with the rapidly growing influence of artificial intelligence, sometimes leads to thoughts similar to those the Luddites might have had in customer communication. However, this is not about humans versus machines, but about the team of humans with machines. When well-orchestrated, they form a fascinating unit capable of tackling nearly all challenges in today’s communication and revolutionizing the Total Experience. **Do you remember who the Luddites were?

To advance Total Experience with artificial intelligence and create new challenges, collaboration between humans and machines is extremely important. However, several factors contribute to this advancement. What these factors are and why they are so important will be discussed in the following chapters.

The primary duty of a contact center has always been, and remains, to create connections and deliver an excellent customer experience. Looking back, this was true in a double sense, as it was mainly airlines that experienced a boom in the 1970s and desperately sought ways to manage the increased booking volume. With Galaxy, an automatic call distribution (ACD) system developed by Rockwell International for Continental Airlines in 1973, machines took over an important part of communication for the first time. They ensured that calls were automatically routed to the next available agent.

The introduction of automatic call distribution coincided with the spread of touch-tone telephones, but the communication channel was limited to the phone and thus to voice. That has fundamentally changed. Today, contact centers must handle all channels, from voice, email, and fax to social media messengers. This can lead to chaotic situations, but technology helps us here as well.

First, it must be ensured that all channels are "ready to receive." This means that whether contact is made by phone, chatbot, WhatsApp, or email, the system should ideally know who is making the contact and how to handle it.

Let's first look at the traditional phone route, but in a version that already uses conversational AI as a gatekeeper. Simply put, this means that AI-powered voice bots answer the call-in natural language, translate the caller's speech into text files, match this with trained data or a CRM using generative AI, and output the result back in natural language.

Conversational AI, especially speech recognition combined with biometric data, can then be used for any necessary authentication. And all of this happens fully automatically and with the highest security standards. For example, during the conversation, the security system is interposed for authentication. After the system's verification, the dialogue is resumed.

What follows the dialogue is crucial for the performance of a contact center. The bot must structure the collected information, prepare it, and pass it to the responsible employee via attribute-based routing. Not only is the next available agent sought, but the specialist who can handle the case conclusively is connected based on a complex set of rules.

For humans and machines to work reliably, they depend on data. In the simplest case, it is the phone number that is recognized and used to display the correct data set from the CRM system. However, this no longer meets the expectations that contact centers must satisfy in the age of automation and artificial intelligence. But for conversational and generative AI to function, they need data. Lots of data! And this data must be provided in various ways.

We distinguish between two forms of data: public and private. For example, if a bot is to provide information about a destination country for an airline, it can use public data from ChatGPT. In the above-described dialogue, however, the bot needs internal data, which is also sensitive. Therefore, the AI must be "fed" and trained internally. It needs to learn standard processes and be prepared with relevant data from databases such as CRM systems and make this data available to both the bot and employees.

For processes to run smoothly, contact center operators need expertise in managing inquiries, automatic speech processing, authentication, connecting CRM systems and other databases, training large language models for generative AI, and many other small and large components necessary for optimal workflow.

It is very unlikely—or very expensive and time-consuming—that these come from a single provider. Therefore, the core of the contact center should offer open interfaces (APIs). This allows specialist providers to quickly integrate their products into the overall framework. ChatGPT is one example.

Ideally, the core system provider already brings a selected set of established partners. This allows individual solutions for outstanding performance to be rolled out in the shortest possible time.

The question of whether humans or machines should be used does not arise. Together, they achieve tremendous results. Humans will never be able to process such large amounts of data at the necessary speed as machines can. However, we are still far ahead of them in the areas of emotion and complexity. Ultimately, we are not only living in a digital transformation. We humans must also adapt to technical possibilities, transform our knowledge and way of life to some extent. Then we play in the team with machines in the top league. With the best chances for a great future and an incredible advancement of Total Experience through artificial intelligence.