In recent years, school attendance problems among university students have been on the rise, leading to issues such as course repetition, academic dropout, and social withdrawal. Although most colleges offer counselling services, many students delay seeking help until their problems become severe. Early identification of at-risk students is therefore critical to prevent long-term consequences. This project proposes a smartphone-based system developed using the MERN stack (MongoDB, Express.js, React.js, Node.js) to detect subtle signs of poor attendance and engagement. The system collects daily data through a check-in form that includes sleep duration, mood, energy levels, and class attendance. Using rule-based and trend-based scoring algorithms, the system evaluates each student's status and flags them as Normal, At-Risk, or Critical. To test the system the results, show that the estimation models are accurate in identifying students with potential sleep or engagement problems. The platform includes two dashboards: one for students to track their trends and receive alerts, and one for admins to view student analytics, send messages, and manage risk reports. This tool does not replace medical diagnosis but acts as an early warning system, helping colleges take proactive steps to support student mental health and academic performance. The system is lightweight, cost-effective, and works entirely on smartphones without requiring any extra devices—making it practical for real-world deployment.

Keywords: Early Risk Detection, Sleep and Engagement Analysis, MERN Stack Application