Keya enhanced data quality for master questions in the TheAnonymous-stack/numi-scraper repository by standardizing validation for correct_answers across all question types. Focusing on fill-in-the-blank questions, Keya introduced support for multiple valid answer formats, such as handling responses with or without percentage signs, to reduce ambiguity and improve downstream analytics and scoring. The work involved disciplined JSON data management, data normalization, and input validation, ensuring greater data integrity and reliability for educational content. Through careful version control and collaborative workflow, Keya’s contributions addressed nuanced data challenges, resulting in a more robust and flexible dataset for educational applications.