The accurate prediction of consumer demand is of paramount importance for the effective management of supply chains. Demand prediction can play a pivotal role in optimizing supply chain management by allowing companies to anticipate future consumer needs more effectively. This study focuses on the integration of decision support systems (DSS) with recurrent neural networks (RNNs) and long-term memory networks (LSTMs) to enhance the accuracy of forecasts. These advanced AI and machine learning techniques have been demonstrated to outperform traditional methods by their ability to capture complex, non-linear patterns in temporal data. This, in turn, has the potential to improve inventory management and reduce costs. Integrating DSS with LSTM models offers several benefits, including the automation of forecast-based decision-making, the optimization of inventory levels, and the reduction of human intervention, thereby speeding up the decision-making process. The results of the BASE2 analysis indicate a root mean square error (MSE) of 0.01, an absolute mean error (MAE) of 0.07, and a coefficient of determination (R-squared) of 0.99, which collectively demonstrate a robust model performance. The current demand is 303.96, with a predicted demand of 297.92. Although the estimated increase in future demand is -1.99%, the DSS recommends maintaining current production levels due to the satisfactory performance of the model. This study presents a conceptual framework for intelligent decision-making based on industrial big data technology, offering valuable insights and research avenues to address major challenges and guide future research in this field.