Development and Implementation of a Hybrid Rule Engine for Nutritional and Lifestyle Recommendations

Increased awareness of health and longevity, coupled with the rapid growth of digital health information, has created a demand for intelligent tools that are capable of taking raw data and turning it into individual, actionable recommendations. This thesis outlines the design and construction of a rule-based nutrition and lifestyle consultancy recommendation system. The architecture is decoupled, separating domain knowledge captured in logical form in JSON from the Python-written analytical center. This offers flexibility, transparency, and maintainability, enabling health professionals to vary recommendations without modifying the source code of the system. The backend, developed with FastAPI, talks to Firebase, the central repository for user details and authentication, in real time. The recommendation process is an organized, three-phase pipeline involving data collection and preprocessing, dynamic rule-based analysis and personalized message generation. A specially developed graphical user interface allows experts to manage the knowledge base in a secure way through syntax checking and version control. Then, the system was enhanced with sleep recommendations, demonstrating the modularity of the design and the capability to integrate additional sides of life, such as physical activity. The final API service facilitates secure, real-time communication with the accompanying mobile application, enabling the engine to serve as a scalable, explainable backend component for personal well-being.