Digital banking has transformed the financial industry, offering unparalleled convenience to customers while introducing heightened risks of digital fraud. This paper presents a robust banking management system aimed at enhancing security through real-time fraud detection and prevention. Developed using Java, Swing for an intuitive user interface, and MySQL for efficient database management, the system leverages rule-based algorithms to identify suspicious transaction patterns, such as more than five transactions or cumulative amounts exceeding ₹50,000 within a one-minute window. Employing a sliding window algorithm, the system ensures real-time monitoring, minimizes false positives, and provides timely alerts for potential threats. The research underscores the balance between robust security and a seamless user experience, offering insights into combating financial fraud in an increasingly digital-first ecosystem.