Mobile devices and smartphones have the potential to be powerful and complex systems. Their ever-increasing popularity has created a need for more robust and personal applications. This paper proposes a personalization system to modify phone's state according to its owner routine and behavior, called Smart Profile. This system was implemented and tested with two modules, a calendar and a location module. They personalize one smartphone aspect, but both aspect and modules can be generalized. Tests performed with data collected from 16 users showed that the system could predict user behavior with elevated accuracy using inferring context data. This paper distinguishes from previous research work in the area by using calendar and geographical location to infer context data, and providing a generalizable personalization model.