In this paper I introduce “monoBeat,” which is a platform that utilizes machine learning to identify heart murmurs in heart sound recordings. This platform provides a cost accessible solution, for detecting diseases at an early stage especially in regions with limited resources. I conducted a comparison between two models; Convolutional Neural Network (CNN) and Long short term memory neural network (LSTM). I used Mel coefficients (MFCCs) and Mel Spectrograms as input features. The results indicate that the CNN model, combined with MFCCs achieved the accuracy; It is the choice for integration into the monoBeat platform. The paper emphasizes how monoBeat has the potential to enhance healthcare accessibility and minimize delayed detection of heart conditions.