HEIDI – A Conversational AI Assistant for Enhancing Guest Experience at Hotel NI-MO Zurich
A customised AI chatbot named “Heidi” was developed for Hotel NI-MO to provide guests with instant answers to common questions, local recommendations, and real-time information 24/7. Built on Google Dialogflow and integrated with live data APIs, it bridges the service gap outside staffed hours.
Amit Jagubhai Jethva, 2025
Art der Arbeit Bachelor Thesis
Auftraggebende Hotel NI-MO
Betreuende Dozierende Jüngling, Stephan
Views: 5
Hotel NI-MO operates with a self-check-in system and limited reception hours (07:00 – 14:00). While efficient, this model leaves guests without direct assistance for much of the day. Frequently asked questions, such as check-in times, breakfast hours, Wi-Fi access, and travel directions, are currently handled via email or phone, creating repetitive work for staff and delaying guest responses.
The project began with requirements gathering in close collaboration with the hotel owner. A chatbot architecture was designed using Google Dialogflow ES for natural language understanding, supported by a Python Flask webhook for dynamic content. Integrations were implemented with APIs such as Google Places, Google Directions, and OpenWeatherMap. Iterative testing with hotel staff ensured accurate responses and usability. Quantitative evaluation of intent recognition accuracy and qualitative feedback were used to refine the system before final handover.
The project delivered “Heidi”, a fully functional AI-powered chatbot tailored to Hotel NI-MO’s operational needs and guest service standards. Heidi automates responses to frequently asked questions—such as check-in times, breakfast hours, and Wi-Fi details—and delivers dynamic, real-time information including travel times, directions, and weather updates through integrations with Google Places, Google Directions, and OpenWeatherMap APIs. Guests can also be guided to a structured feedback form, enabling the hotel to collect input consistently and efficiently. It offers 24/7 availability without additional staffing, reducing repetitive tasks for employees and allowing them to focus on personalised service. By delivering fast, consistent, and engaging interactions, Heidi enhances guest satisfaction and strengthens Hotel NI-MO’s boutique, innovative, tech-forward brand identity. The solution is designed for future growth, with potential for multilingual support, omnichannel communication, and seamless integration with the hotel’s property management system for enhanced service delivery.
Studiengang: Business Information Technology (Bachelor)
Keywords Chatbot, Google Dialogflow ES, Natural Language Understanding (NLU), Hotel Automation, Self-Check-In, Frequently Asked Questions (FAQ) Automation, Digital Concierge
Vertraulichkeit: öffentlich