The current growth in population density has increased the demand for potable water and aggravated water scarcity in reservoirs, especially in urban and metropolitan areas. This problem, coupled with other aspects such as climate change and rain irregularity, has proven that current water management techniques, which rely on human operator knowledge, are insufficient and inefficient. Observing the development of Data Science and Machine Learning techniques, these intelligent algorithms seem to be an interesting approach that could support water management operations by predicting water usage depending on several factors. A first step involves characterizing user consumption behaviors by applying feature extraction and data visualization techniques to discover similarities and common patterns among users.