The construction industry is one of the prominent sectors for not only economic growth but also environmental and social growth. Yet, with modernization, the construction operations have become vulnerable. A significant issue in construction operations is ensuring supplier reliability which directly influences the enhancement of supply chain efficiency and sustainability. This research addresses the issue of precisely forecasting supplier dependability via the use of Artificial Neural Networks (ANNs) which is an algorithm of machine learning. A multilayer perceptron (MLP) model was specifically built to categorize supplier reliability into three classifications which are low, neutral, and high. A synthetic dataset consisting of 101 samples with 10 variables relevant to supplier performance was used for model development and evaluation. The dataset was divided into training (70%) and testing (30%) sections, with feature normalization applied to facilitate successful model convergence. The ANN model, designed with two hidden layers, was trained via a stochastic gradient descent optimizer. The model’s efficacy was assessed using criteria including precision, recall, F1-score, and AUC. The results exhibited increased classification accuracy for all categories, with precision values of 89.74%, 91.39%, and 92.22% for the low, neutral, and high classes, respectively. The recall values were 99.06%, 85.86%, and 87.34%, while the F1 scores varied from 88.54% to 94.17%. The model’s area under the curve (AUC) ratings, above 0.96 for all classes, indicated its superior discriminatory skills. Visualizations, including receiver operating characteristics (ROC) curves, lift charts, gain charts, and pseudo-probability plots, further confirmed the model’s success in evaluating and prioritizing reputable providers. These results underline the potential of ANN-based techniques in strengthening decision-making processes in supply chain management (SCM), opening the door for more sustainable and efficient supplier operations.