This paper introduces a medical suggestion system that effectively aids in the diagnosis and prognosis of heart disease. In order to enhance cardiac disease prediction and medicine use across electronic health environments, fast Fourier transformation is used in conjunction with ensemble learning approaches. The suggested system's accuracy and applicability are assessed in relation to the current heart disease prediction methods. The suggested system helps with cardiovascular risk prediction and recommendation procedures and builds on the existing data mining approaches. It is presumable that hybrid classification approaches will be the foundation for the coupling of quick Fourier transformations with machine learning models. It is presumable that the ensemble model supports accurate prediction and the medical recommendation process, and the use of quick Fourier transformation facilitates time series analysis of the patient's data. Additionally, it is assumed that the input dataset for the proposed system is devoid of noise and missing values.