At Google, I supported LLM Safety and Search Quality programs by creating, curating, and validating datasets for AI model evaluation. I reviewed text, image, and video content against Trust & Safety policies and helped improve evaluation data quality.
One sexually explicit-content dataset I curated was adopted for model evaluation. I also analyzed the TrustNet workflow, identified more than 20 workflow and UX gaps, and documented recommendations for product and operations stakeholders.
I built operational dashboards and reporting workflows that improved leadership visibility by 20%, and contributed recommendations that reduced manual effort by 25%. I also developed a Power BI dashboard for NCR ride-booking data using Power Query and DAX.

