In recent years, smart educational technologies have advanced rapidly and digital learning tools have become widely adopted. However, in specialized and knowledge intensive domains such as Traditional Chinese Medicine (TCM), instructional systems that simultaneously support interactivity, real time feedback, and personalized learning remain limited. To address this gap, this study develops an interactive TCM instructional system that integrates a locally deployed large language model with multimodal sensing technologies to enhance learner engagement, comprehension, and overall learning experience. The instructional content is grounded in the textbook TCM Healthcare Classroom for Everyone, and a Retrieval Augmented Generation (RAG) framework is incorporated so that system responses maintain semantic relevance, factual accuracy, and contextual consistency when answering learners’ question.