<p>Cardiovascular disease is a major global health concern since number of deaths projected by heart disease may reach up to 22.2 million in 2030. However, diagnosis of cardiovascular disease is a time-consuming task by the healthcare practitioners. Traditional methods of diagnosing CVDs are slow and costly and often require various tests and examinations, such as blood tests, electrocardiograph and blood pressure measurements. Hence, there is need to automate the cardiovascular disease detection using machine learning approach. This paper aims to leverage recent advancements in artificial intelligence and data analytics to create a new automated process for early detection of heart disease using machine learning. This research work presents the novel automated process to detect cardiovascular disease by using stacking ensemble method with pipeline machine learning. Proposed stacking ensemble meta-learner (SEML) model is implemented for early detection of heart disease using logistic regression as meta-learner. The method of combining pipeline machine learning with stacking meta-learner improves the efficiency and effectiveness by automatic handling of data transformation, feature scaling, allowing hyperparameter to run all over estimators and crucial preprocessing task. Diverse base classifiers such as logistic regression, SVM, KNN, NB and XGBoost are trained and merged with meta-learner to produce the prediction. The SEML model achieved 91.80% accuracy in predicting cardiovascular disease. The ROC curve for multiple base classifiers shows AUC scores ranging from 0.75 to 0.95, indicating good model performance. F1 score, precision, recall, MCC score and log loss metrics are used to validate the robustness of the model. Experimental findings reveal that the stacking meta-learner with pipeline approach significantly improved the cardiovascular disease prediction, thereby helping medical professionals or any individual at risk for facilitating early interventions, prevention and treatment plan.</p>