Heart disease is a major global health issue that has an impact on rates of morbidity and mortality everywhere. Even with advancements in diagnostic processes, traditional methods struggle with efficiency, scalability, and accuracy. To enhance the prediction of heart frailer, this study proposed a multi-voting ensemble Machine Learning (ML) model that utilizes a several ML techniques as base models such as Random Forest (RF), Logistic Regression (LR), Support Vector Machine (SVM), Naïve Bayas (NB) and Multiplayer perceptron (MLP). To maximize performance like accuracy, sensitivity, and specificity, the proposed method uses the preprocessing techniques including filling missing values, data scaling, and managing categories variables. Utilizing a heart disease dataset containing 918 patient records, the model surpassed the conventional classifiers with a perdition accuracy of 94.13%. This model proposed a useful method for enhancing prompt medication and better patient outcomes. The result shows that the ensemble learning is useful in medical decision-making and in development of diagnostic tools.