Tahir contributed to the WATonomous/humanoid and WATonomous/wato_monorepo repositories by developing a CPU-based Gazebo simulation workflow and an MPC-driven autonomous driving framework. He removed GPU dependencies from Docker configurations, enabling scalable, hardware-agnostic simulation and continuous integration. In wato_monorepo, Tahir integrated Model Predictive Control with CARLA and CasADi, linking trajectory planning to control execution for reliable end-to-end testing. He also stabilized configuration management by implementing a shell-based reset mechanism, improving startup reliability and incident remediation. His work, primarily in Python, Shell, and Docker, demonstrated a strong grasp of robotics simulation, configuration management, and robust deployment practices within complex systems.