AI-Driven Chatbot Streamlines Kärcher’s Customer and Service Support
About Kärcher
Kärcher, a global leader in cleaning technology, offers a wide range of products for both personal and industrial use. With a robust global presence, Kärcher operates over 50,000 service stations across more than 70 countries and collaborates with an even larger number of dealers worldwide.
With over 30,000 products in its portfolio – each potentially accompanied by multiple extensive service manuals – it can be challenging to quickly find the right information, especially for new service technicians. Following a successful IT collaboration with Aixia, Kärcher engaged in discussions on utilizing AI to streamline access to technical details.
\ Challenge & Solution
Multilingual, accurate support simplifies technicians’ work
Together with Kärcher, Aixia developed an AI-driven chatbot capable of interpreting technicians’ questions and providing answers directly from service manuals, in the user’s preferred language. The aim was to enable technicians and customer service representatives to access accurate information quickly without searching through lengthy documents. The chatbot is specifically trained on Kärcher’s products, allowing it to answer questions related to their offerings effectively.
To build the chatbot, Aixia employed a combination of Natural Language Processing (NLP) and machine learning. This solution is designed to interpret and understand technical questions, even when posed in various languages. The AI uploaded and interpreted Manuals and technical documents, which now provides users with precise, concise answers. To ensure high-quality responses, Aixia also implemented a feedback function that allows technicians to rate answers, enabling continuous improvement of the AI model through real-world user interactions.
A unique challenge lies in handling minor but important differences between similar product models, where even a small detail in the model name could imply a different technical solution. Aixia, therefore, developed advanced filtering and search algorithms that help the chatbot distinguish and provide correct information, even with subtle differences. This ensures that technicians receive the most relevant information for the specific machine they are working on.
“To systematically build a robust data structure, we have also focused extensively on prompt engineering and nuances in how users phrase their questions,” says Simon Janeck, Head of R&D at Aixia and project lead.
\ Result
Technicians Receive Direct Answers – Enhancing Service Efficiency Globally
Now, technicians can quickly receive answers to complex questions, locate the right manuals, and access technical support in real time. This has increased efficiency and freed up staff time, allowing them to focus on customer relations and upselling. The solution has been implemented for ten of Kärcher’s most common machine types, with plans for continued expansion.
Aixia developed a sophisticated chatbot based on Retrieval-Augmented Generation (RAG) technology, built on the Microsoft Azure platform and OpenAI’s language models. The RAG method enables the chatbot to combine pre-trained AI with real-time searches within Kärcher’s databases, providing precise and contextually relevant responses. The solution includes vectorization of Kärcher’s extensive documentation, multilingual support to bridge language barriers, and dynamic information updates from Kärcher’s website. An integrated feedback mechanism allows users to rate responses, ensuring continuous improvement of the AI model.
Kärcher is operating in over 70 countries.
Kärcher is a family-owned, global company founded in 1935, specializing in high-quality cleaning solutions for private and industrial applications. The company has a strong international presence with over 13,000 employees in 72 countries and 50,000 service stations.
Read more about Kärcher at:
Case Study Summary
Download a one-page PDF summary of the customer story here.
