The growing prevalence of lifestyle-related health conditions demands scalable, personalized fitness guidance that transcends the limitations of generic workout programs. This paper presents an intelligent generative AI-driven framework that leverages Google's Gemini Pro large language model within a Django web application to produce fully customized, multi-day workout plans tailored to individual biometric profiles. The system accepts four userspecific parameters-age, body weight, height, and fitness goal (weight loss, muscle gain, or general fitness)-and constructs an engineered prompt transmitted to the Gemini Pro API via Google's official Python SDK. The model synthesizes domain knowledge across exercise physiology, sports science, and behavioral psychology to generate structured weekly workout plans calibrated to the user's physiological profile. The application follows Django's Model-View-Template architectural pattern with Bootstrap 5 for responsive frontend presentation and SQLite for data persistence. The framework operates as a stateless request-response system, delivering professional-grade fitness recommendations through a browser-based interface within seconds. Experimental evaluation confirms that the system generates contextually appropriate, structured workout plans that account for user-specific parameters, demonstrating the practical feasibility of integrating large language model APIs within conventional web application frameworks for health and wellness applications.
Naidu et al. (Thu,) studied this question.