Worked on the pytorch/executorch repository to refactor operator test infrastructure, focusing on improving padding handling and restructuring tests for critical operations such as AvgPool, Clamp, Clone, Conv1d, and Sub. Leveraged Python and PyTorch to redesign test pipelines, aligning them with ARM backend requirements and reducing flaky test behavior. The approach emphasized maintainability and scalability, laying the foundation for future operator-related testing. By enhancing unit testing practices and updating quantization-related semantics, the work improved test coverage and reliability, supporting faster feedback loops and more robust model serving in production environments while streamlining backend development workflows for ongoing feature growth.