AI-based assistants, such as conversational agents (CAs) and social robots, are becoming increasingly important in healthcare organizations. CAs provide a scalable and cost-effective platform for organizations supporting their employees by retrieving, structuring, and analyzing information to assist work processes. This study targets how knowledge graphs as ontological models manage CA that help improve the patient flow processes and reduce patients’ waiting time in the emergency departments (EDs). We tailored the design thinking (DT) method with modelling workshops employing conceptual modelling (CM) techniques to address these issues. We incorporated a hybrid formal approach of Methontology and Tove methodologies to build design artifacts, develop a goal-oriented interactive conversational system between humans and machines, and support information systems (IS). As a result, this ontology-driven approach contribution helps developers build value-added CAs to facilitate healthcare practitioners and patients. It is helpful for quality care delivery experience and improves bottlenecks in information flow within Eds.