Business success hinges on how companies interact with their customers. No company can afford to provide inadequate care and support. On the contrary, companies that offer fast and precise handling of customer inquiries can distinguish themselves from the competition, build trust in their brand, and retain people in the long run. Our collaboration with Geberit, a leading manufacturer of sanitary technology in Europe, demonstrates how this can be achieved at an entirely new level through the use of generative AI.
What is generative AI?
Generative AI models automatically create content from existing texts, images, and audio files. Thanks to intelligent algorithms and deep learning, this content is hardly distinguishable, if at all, from human-made content. This allows companies to offer their customers personalized user experiences, interact with them automatically, and create and distribute relevant digital content tailored to their target audience. GenAI can also tackle complex tasks by processing vast amounts of data, recognizing patterns, and learning new skills. This technology enables unprecedented gains in productivity. Routine tasks like data preparation, report generation, and database searches can be automated and greatly optimized with suitable models.
The Challenge: One Million Emails
Geberit faced a challenge: every year, one million emails landed in various mailboxes of the customer service department of Geberit’s German distribution company. It was common for inquiries to end up in the wrong departments, leading to significant additional effort.
The Solution: An AI-powered Email Bot
To correct this misdirection, we developed an AI system that automatically assigns emails to the correct departments. This intelligent classification system was trained with a dataset of anonymized customer inquiries and utilizes advanced machine and deep learning methods, including Google’s BERT model.
The Highlight: Automated Response Suggestions with ChatGPT
But the innovation didn’t stop there. The system was further developed to generate automated response emails. ChatGPT is used to create customer-specific suggestions. Customer service agents only need to review the generated emails and can send them directly.
The Result: 70 Percent Better Sorting
The result of this groundbreaking solution speaks for itself: a reduction of misassigned emails by over 70 percent. This not only means significant time savings of almost three full working months but also an optimization of resources. The success of the project is making waves at Geberit: a central mailbox for all inquiries, expansion into other country markets, and even a digital assistant are in the planning.
Customer Service 2.0 – Innovation, Efficiency, Satisfaction
The introduction of GenAI has not only revolutionized Geberit’s customer service but also demonstrates the potential in the targeted application of AI technologies. Intelligent classification of inquiries and automated response generation not only saves resources but also increases customer satisfaction. A pioneering example of how AI is shaping the future of customer service.