In this paper we demonstrate a fully automated approach for discovering and monitoring patterns of daily activities. Discovering patterns of daily activities and tracking them can provide unprecedented opportunities for health monitoring and assisted living applications, especially for elderly and people with memory deficits. In contrast to most previous systems that rely on either pre-selected activities or labeled data for tracking and monitoring, we use an automated approach for activity discovery and recognition. We present a mining method that is able to find natural activity patterns in real life data, as well as variations of such patterns. We will also show how the discovered patterns can be recognized and monitored by our recognition component. In addition, we provide a visualization component to help the care-givers to better understand the activity patterns and their variations. To validate our algorithms, we use the data collected in two smart apartments.