This research aims to develop a predictive model that helps prepare menus based on customer preferences at Café Minapadi, hoping to improve operational efficiency and customer satisfaction. Using rule-association data mining techniques, the study uncovered hidden patterns in extensive transaction data, applying a priori algorithms in datasets to explore menu ordering frequencies and trends. Data analysis includes cleansing, transforming, and selecting features to generate relevant insights. The results found that items such as coffee and chocolate cake were often purchased together, providing an opportunity for menu optimization and special promotions. Evaluation of predictive models shows the possibility of increased accuracy in stock preparation and adjustment of menu offerings, providing significant benefits in business decision-making in the culinary sector.