During their two-month tenure, Tahir developed and stabilized core autonomous driving infrastructure for the WATonomous organization. In the wato_monorepo repository, Tahir introduced a Model Predictive Control framework for autonomous vehicles, integrating Docker, CARLA simulation, and CasADi-based trajectory optimization to enable end-to-end planning and control on CPU-only systems. In the humanoid repository, they enhanced simulation accessibility by removing GPU dependencies from Gazebo’s Docker configuration, streamlining CI workflows and hardware requirements. Tahir also improved deployment reliability by implementing a shell-based configuration reset, preventing startup failures and supporting deterministic recoveries. Their work demonstrated depth in Python, Docker, and configuration management.