This project focuses on developing advanced machine learning models for tactical decision-making for autonomous heavy-duty vehicles. Key objectives include enabling automated decisions for adaptive cruise control and lane change maneuvers in dynamic highway traffic scenarios. We will employ Reinforcement Learning (RL) techniques and Large Language Models (LLMs) in the decision-making framework. The RL models will be trained with a goal of collision avoidance and reaching target with maximum efficiency. LLMs will be fine-tuned to provide recommendations for high-level driving actions. Training and evaluation will be performed on a simulated highway environment, incorporating both the autonomous truck and surrounding vehicles, leveraging realistic car-following models to capture complex traffic dynamics.