Although the use of pedestrian simulation technologies in architecture and urban planning disciplines is not a new concept; the use of 3D simulation software based on machine learning, which gives results similar to human behavior, which is agent-based and of which validity has been scientifically proven, have emerged in recent years. This kind of software allows seeing circulation and emergency evacuation problems that may arise in crowded and complex spaces while still in the design phase. This way, it allows the necessary revisions to be made in the project before the construction begins. The aim of this study, which is carried out in line with national and international standards, is to examine the use of current simulation technologies in the evaluation of normal operation and emergency evacuation scenarios related to buildings through an actual metro station project selected as an example. In order to find station-specific passenger loads on the chosen rail system line, classical manual calculation methods were used, spatial capacity analyzes were carried out by transferring the resulting values to the Massmotion software, and service levels for all floors of the station were found. As a result, it was tested whether the selected metro station would meet the requirements of regular operation and emergency evacuation scenarios or not, and the results were given within the scope of the study.