In this thesis, I will investigate the development of A Large Language Model (LLM)-based conversational interface to support patient triage in healthcare settings. This thesis aims to create a self-service solution that simplifies symptom assessment for patients. The interaction with the self-service system is enabled by the natural language processing capabilities of LLMs. A human-centered design process is used to ensure that the interaction concept of the LLM-based conversational interface meets the needs of a diverse patient group. Observations, including the evaluation of an existing self-service Check-in terminal in the outpatient department of the Charité Hospital, have provided insights that shape the design process and highlight current limitations. In addition, a literature review has given me a comprehensive overview of current research trends and LLM applications within healthcare. Furthermore, I looked at existing triage chatbots that use hard-coded questions and answers, and found that they often lack flexibility and do not allow patients to ask follow-up questions. This work contributes to human-centered-design research by exploring ways to integrate LLM-based tools into healthcare settings.