Abstract: In the new context of deepening integration between science-and-technology (S&T) innovation and industrial innovation, higher demands are placed on industrial intelligence services. Guided by stage-specific innovation needs and enabled by large language models, deep learning, and related AI techniques, these services must reconstruct a data- and AI-driven decision-support ecosystem, optimize coordination across the innovation, industry, and service chains, and enhance the efficiency and precision of support for industrial technological innovation. Based on the internal logic of interactive integration among the industrial chain, the innovation chain, and the service chain, this study systematically examines the demand and supply of industrial intelligence services and constructs an innovation-demand–driven intelligent service system for industrial intelligence. Oriented toward stage-specific innovation demands, the proposed system mechanistically integrates a data foundation, data-intelligence analysis, and productized delivery, promotes the coordinated coupling of the three chains, and establishes service pathways for knowledge discovery, association analysis, monitoring and early warning, as well as forecasting and evaluation. Through these mechanisms, the system enables a shift in intelligence provision from passive response to proactive service.