To address the challenges faced by traditional learners in acupuncture techniques and clinical applications, such as low efficiency and difficulties in skill transmission, and to eliminate the drawbacks of relying on extensive experience-based teaching, this paper proposes an intelligent acupuncture learning and Q&A system based on deep learning. The system is designed using a layered architecture that includes the data persistence layer, semantic layer, business function layer, and front-end presentation layer. It integrates advanced natural language processing (NLP) technologies from recent years and incorporates a long short-term memory (LSTM) network for intent recognition, enabling question-answer matching for users. This enhances the accuracy of acupuncture-related knowledge extraction and supports interactive learning. Experimental results show that the system performs exceptionally well in knowledge retrieval accuracy, effectively improving learners’ efficiency in acquiring acupuncture knowledge. It provides strong support for the modernization and intelligent development of traditional Chinese acupuncture.