Among the increasing number of health issues related to aging and the growing elderly population, there is an urgent need for new ways to improve well-being among older adults. Recognizing the complex connection between mental and physical well-being emphasizes the importance of comprehensive approaches that address health for the elderly population. In the context of the DAISy project, this paper introduces an architectural framework for a proactive system designed to promote healthy behaviors among the elderly population. Leveraging wearable device sensor data, the system collects real-time user information to generate personalized interventions adapted to individual needs and preferences. At the core of the system’s functionality is the use of advanced generative AI, leveraging Large Language Models to produce customized interactions based on user data insights. By combining sensor data analysis with AI-driven text generation, the system aims to deliver adaptive and personalized experiences, thus addressing the complex challenges elderly individuals face in maintaining their health and quality of life.