Worked on the apache/skywalking repository to enhance profiling and trace data query correctness, performance, and reliability. Addressed SQL logic issues in profiling and trace DAOs by refining filtering, correcting IN bindings, and improving merged-table logic. Moved taskId filtering into the storage layer for AsyncProfilerTaskLog queries, reducing application overhead and improving latency. Optimized trace sorting from O(N^2) to O(N) using pre-indexed data structures, which accelerated UI trace rendering. Expanded unit test coverage for JDBC and Elasticsearch profiling queries, increasing regression detection. Removed deprecated BanyanDB GroupBy.field_name, reducing technical debt. Utilized Java, SQL, and Elasticsearch throughout these improvements.