Voice and chat automation for a major network operator

CreaLog solutions deployed across Europe

As early as the end of 2021, the first new-generation voice and chat bots went live for this customer. They form the foundation for a new generation of AI-based omnichannel self-service with case-closing transactions across all channels. Step by step, more and more processes were subsequently mapped and automated on the self-service portal.

Always with the goal of high customer satisfaction, an optimal customer experience, and cost reduction through automation.

CreaLog Managing Director Michael Kloos is pleased to welcome another well-known client from the telco sector, seeing it as further confirmation of his company's strategy:

"The need for telecommunications providers to answer customer inquiries intelligently and automatically at any time did not start with the COVID-19 pandemic. Supported by Natural Language Understanding (NLU) based on artificial intelligence and machine learning, self-service concepts are a proven means not only of achieving high quality in telephone-based customer service, but also of ensuring it around the clock across all other voice and text channels. The successful deployment of self-service applications at numerous telcos across Europe and Africa shows that our solutions are exactly in step with the times."

The mix makes the difference: on-premise and from the cloud

CreaLog built the extensive system on the CreaLog Service Delivery Platform 7 as a hybrid solution, combining AI-based on-premise components with cloud-based solutions. This allows the telecommunications provider to draw on the best-fit technology from both worlds, which proves to be a clear advantage, particularly during migration.

Intent recognition: understanding what the customer really wants

At a telco, the topics and reasons for contact are more than diverse: they range from technical issues and contracts to invoices and products, and from contract terms to tariffs. A key part of the requirements specification was accordingly the system's ability to interpret the terms and concerns expressed by the customer as accurately as possible, based on semantic analysis.

This is achieved through state-of-the-art Natural Language Understanding (NLU) technology, complemented in the voice channel by speech recognition (ASR) and text-to-speech (TTS) output. This allows customers to freely express their requests and concerns in their own words and interact naturally with the ServiceBots. The entire self-service infrastructure is fully integrated into the service center environment, the live chat, and the telecommunications provider's back-end systems, and also accesses information from third-party systems via an API interface.

Self-service today: case-closing across all channels

A key factor in the solution's strong service orientation: authenticated telco customers log in and resolve their concerns through a voice or text dialogue with one of the many specialized bots - and do so with full case resolution. This allows customers to independently update their address and contact details, change their available data volume, switch tariffs, book add-on options, and much more, by phone or in chat.

Proactive chat bots for a range of tasks

Here too, the telco's requirements were clear: the new chat bots had to be placed on the website so that users could navigate through the pages while simultaneously communicating with the chat bot in a separate window - including on mobile devices. The solution even enables proactive communication with the customer when they are inactive.

To do this, the chat bot initiates questions or suggestions directly related to the topic of the page the user is currently viewing. Or it points out current offers, tariffs, and services. This also works for customers who prefer messenger services such as WhatsApp, Facebook Messenger, Telegram, Viber, or others. In addition, the ServiceBots' shared knowledge base contains numerous answers to off-topic questions in all required languages. The bot also takes conversation context into account during dialogue, interprets emojis and punctuation, and - depending on the context of the conversation - can play rich media such as images or videos.

The latest-generation VoiceBot

In addition to the NLU technologies already mentioned, the CreaLog VoiceBot uses state-of-the-art speech recognition (ASR, text-to-speech). This allows customers to start every dialogue with the open question: "What can I do for you?" If the intent is not recognized correctly right away, the bot asks targeted follow-up questions to further narrow down the customer's specific concern.

Voice output is delivered via the latest, extremely natural-sounding speech synthesis, which can also read out any text from FAQ databases or other information sources. Through the CreaLog Service Creation environment, the telco is able to quickly and easily deploy new voice- or text-based ServiceBots on its own, or adapt existing bots to changing requirements.