Fresh food shops such as small sandwich shops face operational difficulties due to demand uncertainty and the perishable nature of their inventory. These businesses often experience waste or stockouts that directly affect profitability and customer satisfaction. This study develops a data-driven production planning model for a sandwich shop operating in two shifts daily. The model aims to minimize daily ingredient waste and maximize profits. Using real transactional data on sales and purchases, we propose a mixed-integer linear programming (MILP) model supported by demand forecasting. Results show improved alignment between ingredient supply and actual consumption, leading to enhanced profitability and availability of products.