Vamsi Kandula enhanced the ROCm/rocprofiler-sdk repository by implementing advanced Counter Expression Language features, enabling granular filtering and aggregation of performance data. Using C++ and leveraging skills in API design and system programming, Vamsi introduced Select() filtering by dimension indices and added dimension-based reduction operators such as sum, average, min, and max. This work included updates to documentation, core evaluation logic, and test coverage, supporting more flexible and precise performance analysis. Additionally, in ROCm/rccl, Vamsi addressed NCCL API trace version synchronization, improving cross-component compatibility and stability through a targeted version control update, demonstrating careful, high-quality engineering.