I developed a deep learning system for environmental sound classification using the UrbanSound8K dataset. The model processes audio signals, extracts MFCC features, and applies a CNN architecture to classify sounds such as sirens, dog barks, and engine noise. The system uses a fold-based training strategy (train on folds 1–9, test on fold 10) and includes evaluation metrics, visualization, and real-time prediction on new audio files. This project demonstrates AI applications in audio intelligence and healthcare-related acoustic monitoring systems.