The objective of this article is to improve inventory turnover in a company that imports electrical materials for mining in Lima, which currently faces inadequate management with storage costs of S/. 1,045,596.81 per year and a 16.11% impact on turnover. A model is proposed to increase the inventory turnover rate (IR) through demand forecasting and a continuous review system. Logistics 4.0 technologies, such as machine learning algorithms, especially the Multilayer Perceptron (MLP), are also incorporated to develop an accurate predictive model. This will allow for more efficient management, greater accuracy in records, and better demand and EOQ estimates. The model will be validated in a three-month pilot, divided by product families, projecting an increase in IR from 2.87 to 4.36, order fulfillment (IF) to 94.3%, inventory accuracy (ERI) to 95.8%, and a reduction in MAPE error to 9.67%.