Worked on refining the question dataset for the numi-scraper repository, focusing on enhancing data quality and repository hygiene rather than expanding features. The main contribution involved updating and correcting the JSON dataset to improve the accuracy and completeness of scraped question data, which supports more reliable outputs for downstream analytics and model training. Emphasized careful data management and version control through a focused, traceable commit process. Demonstrated skills in data scraping and JSON manipulation, reinforcing best practices for dataset curation. The work addressed the need for precise and well-maintained data, contributing to the overall reliability of the numi-scraper project.