AI-Driven Large-Scale Intelligence Microservices For Predictive Guest Experience Optimization In Smart Hotels

Abstract

Artificial Intelligence (AI), cloud-native computing, and the Internet of Things (IoT) have rapidly impacted the entire hospitality management landscape. This study presents and discusses the development of a complete large-scale AI-based architecture for microservices that can be used to optimize the guest experience in smart hotels in the future. The development and integration of seven different microservices (Demand Forecasting, Personalization, Dynamic Pricing, Sentiment Analysis, Predictive Maintenance, Staff Allocation, Anomaly Detection) in a modular manner, along with a containerized deployment pipeline. The data used for experimental validation comprised 847,320 guest interaction data from three five-star smart hotel properties over 24 months. The hybrid ensemble model outperformed the remaining models with a predictive accuracy of 95.3%, an F1 score of 94.3%, and a root-mean-square error (RMSE) of 2.18. Post-deployment results showed an overall increase in guest satisfaction scores of 37.7%, an average check-in time reduction of 66.9%, and an increase of 39.8% in Revenue Per Available Room (RevPAR). The results confirm the potential of distributed AI microservices as a scalable and manageable approach to intelligent and smart hotel operations and their impact on smart hotel design and service delivery, focusing on the needs of the guests.