The general idea of the AMBER/Child Alert System is that by broadcasting and distributing information about a missing child to the community, the public's involvement can trigger critical feedback that would have otherwise been ignored. This feedback, in several cases, can be proved decisive in finding the missing child. Despite the efforts at country and European level to effectively address the issue of missing children, a number of key challenges remain open, including lack of location-focused distribution of alerts, insufficient capture and diffusion of information, and lack of a mechanism that uses and merges all available sources of information. The aim of this paper is to present a novel approach for handling such challenges through a data analytics platform and a mobile application available to all citizens. Using the active research fields of human mobility pattern analysis and machine learning, we show that missing children investigations, as well as search and rescue operations, can be actively supported and enhanced when multiple data sources are combined and analyzed.