The objective of this project is to develop a game playing agent for Super Mario Kart using different Deep Learning techniques. Convolutional Neural Networks and Reinforcement Learning techniques are used to build an agent that will be able to navigate through certain tracks in the game and use the special powers provided to the characters to eventually win the race.