Abstract: Technology transfer serves as a vital bridge between scientific innovation and industrial development, and plays an essential role in advancing the deep integration of innovation and industrial chains. In recent years, both the scale of technology transfer and the level of patent commercialization in Chinese universities and research institutes have continued to improve. Nevertheless, significant challenges remain in linking research outputs with industrial needs, including inefficient demand-supply alignment, limited discoverability of relevant achievements, imprecise matching, and insufficient service efficiency. In the context of technology transfer at new-type research institutions, conventional practices still rely heavily on human experience, making it difficult for enterprises to identify suitable technologies and expert teams, increasing the cost of communication and demand interpretation, and leaving expert selection and resource allocation without adequate intelligent support. To address these issues, this study develops the Tianji Xing system based on the technology transfer practices of the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences. By integrating large language models, knowledge graphs, and talent profiling, the system supports the organization of multi-source scientific and technological resources, enterprise demand understanding, achievement retrieval and recommendation, and expert team matching. The system has been deployed in technology transfer practice and has facilitated the intelligent upgrading of achievement presentation, demand analysis, and institute-enterprise matchmaking. It has demonstrated promising application performance in improving achievement retrieval efficiency, expert matching capability, and the overall level of technology transfer services.