Post-harvest fruit losses in Brazil can reach up to 40%, with inadequacies in the cold chain being one of the primary causes. This study proposes the development of a neuro-fuzzy model to predict the pulp temperature of mangoes in rapid cooling chambers, aiming to enhance the efficiency of the cooling process. The experiment was conducted on a commercial mango farm in Petrolina, Pernambuco. The results demonstrated that the neuro-fuzzy model can accurately estimate the pulp temperature of mangoes (R² = 0.98), thereby aiding decision-making related to optimal rapid cooling times. Implementing this model could significantly reduce post-harvest losses and help ensure the quality of the final product.