Universities face challenges in large-scale data training and data privacy protection when using large models.An Artificial Intelligence Large Language Model called SynHuaAI which was constructed for Huashang Smart Campus was proposed in this paper.SynHuaAI has the characteristics of data privacy protection, efficiency, generalization, and accuracy in answering questions.The data privacy protection of SynHuaAI benefits from its local deployment and large-scale local data training mode.The efficiency and accuracy of SynHuaAI benefit from the parameter fine-tuning technology and semantic analysis technology of the knowledge base for large models adopted in this article.Generalization of SynHuaAI is achieved by supporting access to the ChatGPT-4 API interface on the cloud or other open-source large model interfaces. The deployment process and typical application demonstration scenarios of SynHuaAI was also described in detail.The experimental results of the prototype system show that the SynHuaAI large model can accurately answer questions and improve the work efficiency of university teachers and students.