AI Assistant for Physiology and Physical Well-Being of Students

Students in higher education often contend with erratic sleep, little exercise, poorly balanced diets and mounting academic pressure, yet most campuses provide no continuous, individualised support for these daily habits. This paper describes the design of an artificial-intelligence assistant that promotes the physiological and physical well-being of students. The system collects brief daily self-reports and optional wearable readings, turns them into six sub-scores tied to published health guidelines, and combines these into a transparent wellness index from 0 to 100. A hybrid decision layer, made up of explicit rules, a machine-learning risk classifier and a retrieval-grounded conversational module, generates tailored advice, while a safety guard forwards signs of serious distress to human counsellors rather than replying automatically. We present the layered architecture, the scoring equations, the recommendation logic, the data model, the privacy controls and a protocol for a pilot evaluation. The framework is modular, explainable and cheap to deploy in an engineering-college setting.