Machine learning and artificial intelligence have been used in various fields, including academic and industries, to make driving safer, more reliable, and improve performance. The objective of this research was to create an agent framework capable of driving a vehicle autonomously in a simulated world. The simulation environment was designed to receive control actions from both agents and humans while providing real-time updates on the state of the vehicle through in-vehicle sensors. The agent framework was developed to learn from past experiences and improve its driving skills. The results showed that the agent can drive the vehicle smoothly and safely. The data collected from the simulation verified the outcomes, exhibiting stable driving behavior and moving speed. This research has potential to accelerate the development of the automotive industry by increasing driving safety without compromising on performance.