The aggressive actions seriously jeopardize one’s quality of life and personal safety. To curb the hostile conduct, several different strategies have been tried, such as installing and keeping up surveillance cameras. It will be extremely relevant if the monitoring systems can recognize violent conduct automatically and immediately send warning or alarm signals to the authorities. New researchers are being drawn to the busy study fields of computer vision and image processing for the identification of aggressive and aberrant conduct. Using MobileNetV2 for violence identification in video streams, this research suggests a real-time violence detection and alarm system. With the help of Telegram Bot for real-time alerts and Cloud Firestore for data management and storage, the system offers public safety a scalable, precise, and flexible solution.