Abstract: Faced with challenges such as weak fire supervision forces, outdated traditional supervision models, difficulties in multi-source data fusion, and insufficient regulatory efficiency, the Tianjin Fire and Rescue Brigade actively responded to the requirements of the national development of new quality productivity and big data strategy, with the goal of creating new quality combat effectiveness, and developed a fire risk monitoring and early warning system to lead and guide the comprehensive supervision of fire safety in the city. This article systematically elaborates on the establishment of a fine-grained fire risk warning model, a large model in the field of fire protection, and a fire risk traceability model by the Tianjin Fire and Rescue Brigade, relying on the technical strength of the National Supercomputing Center and other technologies, with big data as the foundation and AI intelligent computing as the support. It also integrates multidimensional real-time monitoring data to achieve precise push and penetration of warning information to local, industry, and professional regulatory entities, significantly improving the accuracy of risk identification, the efficiency of regulatory resource allocation, and the closed-loop rate of fire hazard rectification. This provides strong support for building a scientific and efficient comprehensive fire safety supervision system and promoting the transformation of governance to "pre-vention" mode.

Key words: fire risk monitoring, intelligent fire warning, big data, fire supervision and management