Traditional stethoscopes involves manual role in the diagnosis of heart, lungs, etc. related abnormalities. This novel work presents the design and implementation of low-cost, prototype, smart stethoscope that records, stores, and shares heart sound signals in real-time. The proposed stethoscope converts acoustic signals into electrical signals and extracts relevant features from the audio signal to provide insights for heart functions. The system includes, the acquisition of audio signals from human heart using a high-fidelity electret condenser microphone. And extract the features such as Spectral centroid, Spectral bandwidth and Spectral Roll-off from heart sound. All these features fed to the Support Vector Machine (SVM) for classification. The proposed methodology has been examine using performance metrics: accuracy, precision, recall, and F1-score. The technique adapted results with accuracy 85% in classifying normal and abnormal heart patients, with a precision of 0.82, recall of 0.88, and F1-score of 0.85. This research will facilitate the subjects in monitoring their heart abnormality at home and aiding to the physician for diagnosis. The future work will extend the performance of system for lungs’ diseases and more heart abnormalities with large real-time medical datasets required.