<p>In today’s complex business environment, highlighted by challenges like the COVID-19 outbreak, it is crucial to enhance supply chain resilience and agility. Additionally, the fourth industrial revolution and digital transformation necessitate the integration of advanced technologies into supply chains. Moreover, the importance of sustainability, encompassing environmental and social factors, has become increasingly prominent across various sectors. The dairy industry, particularly in chain stores, stands out as particularly sensitive to adopting these paradigms. In this regard, the main goal of this study is to design a model to evaluate suppliers of dairy products in chain stores, which is done using data-driven methods. Due to the expansion of the available data volume and the higher accuracy of the data compared to the intuition of experts, in this study, a new data-driven approach to evaluate and predict the performance of suppliers is presented. In the first step, according to the collected data, clustering is done with the <i>K</i>-means algorithm, and then, using the random forest algorithm, the evaluation and prediction model of the suppliers’ performance is designed. The random forest algorithm, with an accuracy of 92%, has outperformed the <i>K</i>-nearest neighbors (KNN) and artificial neural network (ANN) algorithms. In order to increase the accuracy of the model, the principal component analysis (PCA) algorithm is used, and by applying this algorithm, the accuracy of the model reached 98%. Additionally, the Shapley additive explanation (SHAP) algorithm was used to conduct a sensitivity analysis of the features influencing supplier evaluation. The findings indicate that delivery speed, product quality, and the supplier’s financial capability have the greatest impact. In summary, the main innovation of this study is the development of a data-driven multi-combination model for supplier evaluation.</p>