The acquisition of real estate presents a challenge when considering personal preferences for this decision. Typically, this decision is based on the proximity to services such as shopping and leisure areas, access to public transportation, and general public services, including hospitals and educational centers. Currently, real estate agents use empirical methods or other properties with similar characteristics to predict a price for those seeking this service. This study aims to design an automated real estate valuation system based on data mining and geospatial information techniques adapted to the urban areas of the Metropolitan District of Quito to generate reliable, objective, and efficient estimates of the value of real estate. To achieve this, we followed the guidelines of the Design Science Research Methodology, which allows us to develop a product and validate its results. From this, we created a prototype that enables real estate price prediction through data mining. Specifically, we used the clustering technique, which utilizes the k-means algorithm to group properties that exhibit specific patterns. We calculated the k-value using the Silhouette method and the Elbow method. The results demonstrate the functionality of the application, which would streamline the aforementioned evaluation process.