The expansion of a Swiss railway company into Italian-speaking regions of Switzerland has presented communication challenges for German-speaking train drivers. This language barrier has caused safety concerns. Consequently, there is a need to teach German-speaking train drivers Italian to improve communication in the workplace. To address this, the company seeks an innovative and flexible solution that can seamlessly integrate into the drivers' daily routines, ensuring effective communication in routine operations and emergencies. A conversational AI chatbot could be the answer.

The bachelor thesis focused on creating a chatbot prototype which was developed using the Voiceflow platform. The prototype underwent two rounds of testing, first with a focus group of clients and then with a focus group of four train drivers. The chatbot was based on one of the Italian lessons the train drivers had previously received in a traditional classroom setting. It features engaging dialogue-based interactions with context-specific scenarios and provides instant feedback, thereby offering a dynamic and practical learning experience.

Feedback showed that the chatbot could improve Italian language skills, with users appreciating its easy-to-use interface and real-time, relevant feedback. Train drivers particularly valued the context of real-life scenarios relevant to their work. Additionally, the flexibility to use the chatbot during breaks and commutes was highlighted as a key advantage. The tool's capacity for continuous updates and scalability makes it a sustainable solution for ongoing language training needs. Suggestions for future enhancements included adding progress tracking and incorporating tolerance for minor errors. Overall, Roberto shows strong potential as an AI-driven educational tool to support language training in the railway industry.