Hazardous materials are commonly found in industrial areas, where various companies use them in the production and manufacturing of household appliances, food, construction materials, automotive mechanics, fishing, etc. Activities in these industrial areas carry a high risk of emergencies involving hazardous materials, which can cause harm to people, infrastructure, and the environment. To address this problem, a predictive model was developed to assess the level of risk associated with the storage and management of hazardous materials in each company within the industrial area. The model development followed the CRISP-DM methodology and was implemented using Python to apply machine learning techniques. Decision trees were chosen due to their ability to effectively visualize and interpret the model and its attributes. Furthermore, the results are presented graphically using the QG IS geo-graphic information system (GIS) software, which provides a visualization of the risk associated with each company.