Air transportation is one of the fundamental elements of the global transportation network and plays an important role in both commercial and passenger transportations. The increasing number of passengers every day is leading airlines to seek more efficient and sustainable solutions in operational processes. The food supply process is a critical factor in maintaining passenger comfort and service quality, especially on long-haul flights. The amount and variety of food offered during the flight requires effective planning in terms of both customer satisfaction and operational efficiency. This research presents a data-driven approach to optimize food loading processes in the airline transportation sector. The main objective of the project is to increase operational efficiency, reduce costs and minimize food waste by predicting the amount of food required for flights. In the study, variables such as flight route, number of passengers, flight duration and total demand are considered and the effects of these factors on food consumption are analyzed. In the model development process, machine learning algorithms were applied using real flight data. The data set taken to train the model includes detailed information about approximately 180,000 flights. This data is divided into two as training and test sets in order to improve the learning ability of the model and increase the prediction accuracy. In order to evaluate the prediction performance of the model, comparisons were made with the real consumption values at the end of the flight. The accuracy rate of the developed model shows that the model has a general prediction ability according to the initial findings. The study has the potential to contribute to the improvement of food loading processes of airline companies with data-driven decisions. Development steps such as expanding the data set, model optimization and error analysis are suggested for higher accuracy rates.