Smart Cities is the future of human habitation, which is evangelized by the Internet of Things (IoT) technology. We study the municipality of Papagos, which is located in the Smart City of Athens, Greece. In Papagos is developed a technical infrastructure, which enable citizens to act as human sensors by exploiting their smartphones to report malfunctions in the municipality infrastructure. Using Citify software app municipality problematic situations annotated and submitted to the system for further processing exploiting crowdsourcing and crowdsensing technology. When a report arrives to the municipality control center the system allocates certain department to serve the problem. Since incidents are served by a certain number of departments with limited resources, the early planning and allocation of a department's resources before the incident emerges is significant. To handle such situations, we incorporate an inference engine model, which is based on a Long Short-Term Memory (LSTM) Neural Network to learn stochastically examples of incidence occurrence. Based on the LSTM classification model the proposed system is able to predict a similar event in the future thus allocates efficiently a municipality department resource before the problem emerges.