Maria Racai enhanced the common-voice/common-voice repository by implementing Romanian language localization features, focusing on supporting regional dialects and accents. She extended the existing data model using TypeScript to incorporate Romanian variants, improving localization accuracy and user interaction for Romanian-speaking users. Her work involved stabilizing variant and country code mappings, which reduced the risk of misclassification and improved downstream data processing. By emphasizing data modeling and localization best practices, Maria ensured the changes were review-ready and traceable. This contribution laid a foundation for future dialect coverage and data quality improvements, demonstrating depth in both technical execution and domain understanding.