Worked on enhancing benchmarking capabilities for FFT and overlap-add (OA) convolution routines in the scipy/scipy repository. Focused on refining parameter sets and clarifying test function names to improve the accuracy and clarity of performance comparisons between fftconvolve and oaconvolve methods. Leveraged Python and numerical analysis skills to standardize benchmarking processes, enabling faster iteration and more reliable measurement of convolution performance. Improved documentation and naming conventions to reduce ambiguity in future benchmark tests, supporting easier onboarding for new contributors. The work emphasized maintainability and reproducibility in benchmarking, contributing to a more robust framework for evaluating signal processing algorithms in SciPy.