Cardiovascular disease (CVD) is a major health concern that impacts significantly to mortality and morbidity throughout the globe. This study adopted a comprehensive approach for predicting CVD using Deep Learning (DL) models. It uses an existing dataset from Kaggle. In the first step, an exploratory data analysis was done on the considered dataset to understand the features and their contributions towards disease prediction. In the second step, three baseline DL algorithms such as Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BiLSTM) were implemented, including a hybrid approach with a fusion of CNN and BiLSTM for CVD prediction. It is observed that the hybrid model gave an accuracy of 98.5% which is higher than the baseline algorithms.