My PhD research aims to develop a general framework for a behavior control architecture that will provide customized interaction between the robot and an elderly individual suffering from mild cognitive impairment (MCI). This framework will enable the robot to learn from past events and to adapt its behavior to the specific needs of the person that it interacts with. For this purpose, models will be created for both verbal and non-verbal communication. The user profile (e.g., personality, cognitive disability level, emotional internal states, preferences) will provide the input based on which the robot will adapt its behavior.