The object of study is the timely prediction of the possibility of non-compliance in terms of timings or financial targets. This research will test the potential of a system capable of generating an alerts for the possibility of non-compliance based on known data at the time of application or at key moments in the project's monitoring. This alert should be substantiated according to the studied variables that are directly involved in this result, in order to support a properly informed decision.   

The project aims to improve incentive management, using machine learning approaches to identify operational and strategic risk levels in the analysis and verification phases of project payment requests.