Abstract: Embodiments herein provide a method and system for a real-time allocation of a preferred room type based on one or more preferences of a customer. Every hotel maintains its own reservation database, the system gathers some of property management system (PMS) attributes of that day from the hotel administrator. The PMS attributes includes customer's booked room type, number of rooms booked, number of days customers will stay. Further, the system verifies all the preferences of the hotel and the different room type that are available for offering to customers on the day of check-in. A Check-in Check-Out model takes the attributes such as customers' reservation details, weather, arrival, and departure patterns of the customers from the transportation data store to predict the check-in and check-out time of the customers using pre-trained machine learning prediction models. A looping technique maps the rooms to the customers according to the preferred room types.