<p>In today’s healthcare, making sure medical equipment is reliable and important supplies are available helps maintain excellent patient care. The study introduces a data-based solution to improve both predictive maintenance and inventory management in medical device supply chains. The research uses a Kaggle dataset that covers equipment usage, maintenance records, inventory levels and orders to emphasize the importance of preprocessing the data by handling missing values, normalizing it and detecting outliers. Random Forest, Long Short-Term Memory (LSTM) and XGBoost algorithms are used together to create a hybrid predictive maintenance model. The model uses operational time, usage cycles, error logs and environmental factors to estimate when equipment might fail. At the same time, the ARIMA model for time-series forecasting is used to predict future demand for inventory using past records. The main novelty is that the research brings together predictive maintenance and inventory forecasting models. The strategy helps hospitals stock more of the spare parts needed for high-risk devices and time their inventory replenishment according to device usage. Because of this, the supply chain could operate more efficiently by avoiding overstocking, reducing sudden purchases and maintaining continuous service. The methodology we suggest helps hospitals save money and makes their supply chains more reliable and flexible. This research shows that machine learning and statistical forecasting could guide smarter, data-based choices in healthcare logistics.</p>