Arrhythmia stands as one of the most significant global health concerns, bearing a substantial burden of mortality worldwide. Timely identification of Arrhythmia holds potential for saving numerous lives on a global scale. Arrhythmia denotes an irregularity in the heart's rhythm, which can manifest as either excessively rapid or slow heartbeats. Not all instances of arrhythmia prove fatal, early detection presents an opportunity to prevent millions of deaths worldwide. The advent of the digital age has revolutionized everyday activities through the integration of Internet of Things (IoT), sensor technologies, and advancements in machine learning (ML) and deep learning (DL). Recent advancements in deep learning models have significantly contributed to the classification of heart diseases by analyzing electrocardiogram (ECG) readings. Convolutional neural networks (CNNs) and Autoencoders have emerged as prominent tools for arrhythmia detection, particularly in the integration of CNN techniques with other methodologies, which augments the predictive capabilities of cardiovascular detection systems. This research paper presents a novel Arrhythmia Prediction Model, naming ARR_PRED_EN_LSTM incorporating Autoencoder architectures along with LSTMs. The proposed model demonstrates superior accuracy in arrhythmia detection from ECG recordings, leveraging cutting-edge technologies for efficient and precise identification of irregular heart rhythms.