PsyFit: Intelligent Platform for Personalized Training and Nutrition Using Artificial Intelligence and the Psycho-Fitness Model.

Abstract

This Master's thesis presents the design, development, and implementation of PsyFit, an intelligent, personalized fitness and nutrition web platform that bridges physical training and psychological well-being. Conventional platforms focus exclusively on physical met rics, whereas PsyFit introduces a dual-assessment framework evaluating physical profiles (including Body Mass Index and metabolic parameters using the USDA Food Data Central database) alongside psychological states (motivation, energy, stress, and behavioral habits). The system utilizes a modern development stack consisting of React.js, Vite, and TypeScript on the frontend, and Node.js, Express.js, and MySQL (via Prisma ORM) on the backend. Furthermore, the platform integrates Google's Gemini 3.5 Flash LLM via the Vercel AI SDK to provide continuous, empathetic wellness chatbot support. While the prototype demon strates immediate feasibility, constructing a fully automated psycho-fitness model remains a direction for future work due to project time constraints. Ultimately, PsyFit establishes a robust foundation for holistic, next-generation wellness platforms.