There is a growing body of multidisciplinary research on how robotic systems can be deployed in education and training by providing personalized tutoring session to the user. Socially Assistive Robotics (SAR) is an efficient tool for educational and health-care purposes. In this work, we present our SAR system for personalized and adaptive cognitive training. More specifically, we present the sequence learning task that provides measures for executive function assessment, which may indicate learning or even behavior disabilities in children. This work outlines the designing and evaluation process of such a system, including data collection and analysis. The long-term goal of this research is to develop interactive machine learning methods towards the design of an adaptive SAR system that provides a personalized training session by adjusting the session parameters and the robot’s behavior to maximize user engagement and performance.