Worked on the promptfoo/promptfoo repository to enhance the reliability and scalability of evaluation workflows by addressing issues with large payload transmission. Focused on backend and API development using TypeScript, the solution introduced server-side prompt stripping to manage payload sizes, enabling efficient serialization and client delivery of extensive evaluation tables. The approach prioritized removing the largest prompts first, effectively reducing transfer times and preventing client-side timeouts without compromising evaluation integrity. Emphasized robust data and error handling to ensure end-to-end data flow remained stable under heavy loads. This work improved the overall efficiency of large-scale evaluation pipelines within the project.