Residential fires during cooking activities pose a significant safety concern. In many circumstances, household fires provoke human burns, injuries and sometimes fatalities. Therefore, the development of a comprehensive and effective cooking fire prevention system is crucial. In this paper, an architecture of a smart cooking fire prevention system is presented. The proposed architecture integrates a network of Internet of Thinks sensors, a Fuzzy Inference System, and an alert and emergency generator unit to mitigate cooking-related fires in real time. The network of sensors, which monitor environmental conditions, includes temperature, smoke, gas and proximity sensors. Data from these sensors is transmitted to a central processing unit where a fuzzy inference system assesses the emergency level of a potential fire. The estimated emergency level is transmitted to the system’s alert and emergency generator unit, which decides about the necessary emergency actions, that include residents’ notification through mobile alert messages, fire suppression devices activation, fire alarm activation, automatic dialing of the fire brigade. The proposed architecture allows the continuously real-time monitoring. It aims to detect the possibility of fire and prevent cooking-related fires and the damages they can cause. This research highlights the need for a comprehensive cooking fire prevention system and underscores the importance of integrating advanced technologies like Internet of Things and Fuzzy Logic in developing intelligent fire prevention solutions.