Worked on improving data acquisition reliability in the tensorflow/datasets repository by addressing issues with downloading large files from Google Drive. Implemented a backend fix in Python that extracts the actual download URL from Google Drive confirmation pages, allowing files to be retrieved without triggering virus scan warnings. This solution involved web scraping techniques and enhanced the file handling logic within the downloader module. The update reduced download failures and improved the experience for users accessing Google Drive-hosted datasets, while also lowering support overhead for data ingestion pipelines. The work demonstrated depth in backend development and practical problem-solving for data workflows.