In response to the increasing need of assistance services for people with disabilities and elderly, eAssistance focuses on the development of tools to increase their autonomy and self-sufficiency. In this paper, we present a new eAssistance system based on Ambient Intelligence (AmI) designed to monitor the user’s activities and to improve the self-sufficiency together with the Quality of Life of dependents. The system can be adapted to a wide range of users and easily integrated into their homes or residences. The remarkable novelties of the proposed system are the inference of user’s behaviour patterns with the support of the home automation system, the obtainment of high level conclusions and the possibility of identify derivations and anomalous actions.