Emergency departments often face long waiting times and limited staff availability, leaving patients without adequate information or emotional support. This thesis examines whether a personality-adaptive conversational agent can improve patient experiences in the waiting area by providing reassurance and tailored communication. Building on the ILLUMINATION project’s exploration of Large Language Models in clinical contexts, the thesis explores implementing an Artificial Intelligence (AI) companion that uses real-time text-based personality recognition to adjust its conversational style.
A mixed-methods user study will evaluate the system by comparing AI-predicted personality profiles with participants’ Big Five assessments and by collecting user feedback on satisfaction and perceived reassurance. The findings aim to inform the feasibility and usefulness of personalized AI companions in emergency care and contribute to the design of patient-centered adaptive systems.