In this paper, we focused on designing a system that can guide and train a user in painting an art work. Initially, we aim to develop a system which can guide a user with basic strokes of Korean language calligraphy. The proposed system is implemented in a sequence of three steps. Firstly, we collected the data from the surfaces of the canvas and different orientations of a painting brush. Then based on the data, a relationship is established among the collected parameters by building a machine learning model. Finally, actuators attached with the handle of the brush provides the vibrotactile and force feedback based on the built model. The actuator guides the user in order to paint the required object.