Lifestyle Medicine (LM) has demonstrated significant potential in reducing morbidity and mortality by addressing behavioural risk factors. However, its clinical application remains limited within the Dutch healthcare system, where clinicians face substantial barriers to integrating lifestyle counselling into routine practice. These challenges stem from time-constrained consultations (10-15 minutes) and inadequate systematic training in relevant skills and knowledge. Current training methods, including traditional workshops and online courses, lack either scalability or the ability to develop practical, hands-on competencies, creating an urgent need for innovative training solutions suitable for real-world hospital environments.

This graduation project, commissioned by Leiden University Medical Centre (LUMC), addressed the research question: "How can conversational agents be designed to train healthcare professionals in delivering lifestyle counselling within the 10–15 minute consultation time frame?" The study employed a comprehensive research methodology including literature review and two phases of user research to identify and reframe the core problem. The research revealed that the commonly cited "lack of time" barrier is actually a symptom of a deeper "professional capability gap," encompassing insufficient practical skills, low confidence, and inadequate LM awareness. Focused interviews with clinicians in hospitals further identified a "systemic support gap," characterised by a lack of structured workflows and integrated infrastructure.

Based on these findings, the study developed WellSaid, an AI-driven digital ecosystem designed to train and support clinicians in lifestyle counselling. The solution comprises three integrated components: Consultation Training featuring an AI conversational agent for role-play practice; a Lifestyle Toolkit providing practical resources and support materials; and a Library serving as a centralised hub for accessible, evidence-based information. Expert review by LUMC clinicians confirmed the prototype's desirability and conceptual feasibility, validating that AI-driven conversational training ecosystems can effectively address professional capability gaps. This project demonstrates that AI-based conversational training ecosystems can effectively equip healthcare professionals with practical skills, confidence, and systematic support necessary for integrating lifestyle counselling into routine hospital care, promoting broader LM adoption and supporting more preventive, patient-centered healthcare delivery.