<p>This paper presents the design and implementation of an Edge AI-based Sound Surveillance and Alert Recognition System using an STM32 F401RE microcontroller. The system utilizes an Electret microphone to capture sound data, which is then processed by a lightweight machine learning model tailored for edge devices. A custom-trained Support Vector Machine (SVM) model is employed for real-time classification of sound events such as alarms. sirens. and environmental noises. The addition of a sound level sensor enhances detection reliability and system robustness, particularly in dynamic and noisy environments. Challenges related to sound classification in overlapping noise conditions are discussed along with potential improvements for future system iterations.</p>