Developed and integrated GPU Roofline Analysis capabilities into the intel/pti-gpu repository, enabling performance estimation of GPU kernels through the roofline model. This work involved creating Python scripts and configuration files to automate analysis workflows, as well as updating documentation and the README with clear usage examples to support adoption. Focused on GPU computing and performance analysis, the implementation provides a foundation for data-driven optimization of GPU workloads. By enhancing the unitrace tool with these features, the developer established groundwork for future analytics and feature expansion, leveraging skills in scripting, documentation, and technical communication to facilitate broader GPU performance insights.