During July 2025, Baishali Mullick focused on backend development for the nod-ai/SHARK-Platform repository, addressing a critical bug in the SDXL inference API. She updated Python classes in python_pipe.py, renaming InferenceExecRequest and GenerateService to SDXLInferenceExecRequest and SDXLGenerateService, respectively, to ensure that Stable Diffusion XL model processing used the correct SDXL-specific components. This targeted fix improved the accuracy of inference requests and stabilized the end-to-end workflow, reducing runtime errors and misrouting. By clarifying the codebase structure around SDXL components, Baishali enhanced maintainability and supported future onboarding, demonstrating depth in Python and backend engineering practices.