Abstract: 【Purpose/significance】In response to the new scientific research dynamics brought about by AI4S (Artificial Intelligence for Science), this study explores the application modes, platform systems, and microservice designs of knowledge innovation services, aiming to more effectively support technological and scientific innovation activities.【Method/process】This study proposes an applica⁃ tion mode for knowledge innovation services based on the SECI model and AI. It employs microservice architecture theory and domain-driven design principles to construct the overall framework of the knowledge innovation service platform, with detailed de⁃ signs for the microservices of core business functions.【Result/conclusion】An application mode for knowledge innovation services cen⁃ tered around domain knowledge metadata repositories, collaborative innovation tools, and an academic community hub has been estab⁃ lished. The platform's system functions focus on the implementation of knowledge transformation and innovation. The microservices encompass eight components: user management, knowledge metadata extraction, retrieval, scientific and technological information ser⁃ vices, question-and-answer interaction, personalized recommendations, collaborative project management, and research team manage⁃ ment.【Innovation/limitation】Propose and construct the knowledge innovation service application model that combines the service needs of the research lifecycle with the characteristics of AI driven scientific research transformation. During the design of the plat⁃ form's microservices, there remain areas where the definition of domain events has not been fully considered or optimized.