This research aims to develop a conversation-based system for school building safety inspections. The safety of school buildings requires routine and additional post-disaster inspections. However, the inspections are often complicated due to various timings and sets of checkpoints for different situations. In addition, the traditional paperwork process is not convenient for assessors to add additional images for detailed descriptions. Moreover, managers may face the inefficiency of grasping situations from numerous paper reports. To solve these problems, we developed a chatbot to notify, guide, and assist assessors to complete safety inspections, and a dashboard for managers to consume reports in order to determine whether further assessments or retrofits are required. In this research, we digitalized the process of safety inspections and developed a chatbot to notify and guide the assessors to complete their tasks. When an earthquake occurs, the chatbot notifies the assessor if the intensity scale of the earthquake meets the school buildings safety threshold. The chatbot then guides the assessor to complete the inspection. The assessor may interact with the chatbot for further instructions and upload pictures of damaging. The system collects the reports from all schools, analyzes the data, and displays via the dashboard we designed for the managers which enables the managers to efficiently and correctly consume the reports for making decisions. The conversation-based system provides an effective interface for reducing the inefficiency of school buildings inspections, featuring automatic notifications for assessors, a conversation-based interface for guiding assessors to accomplish inspections, the integration of intuitively concatenating the whole process of inspections, the acceptability of multimedia for lowering possibility of inspection mistakes, and the view of data visualization for decision supporting of the manager.