At Microsoft, I improved product tracking and notification experiences through data analysis, hypothesis testing, ML prototypes, and GPT-based classification. My work increased back-in-stock notifications by 150% and reduced DSAT by 19% for price-drop notifications and 17% for back-in-stock notifications.
I’ve also built large-scale crawling, cleaning, ETL, and data-quality pipelines for millions of products, automotive marketplace data, and internal services. I developed .NET APIs for GPT-powered customer-data auto-labeling in Microsoft Language Studio and created dashboards that helped teams monitor and resolve data anomalies faster.
Earlier, I trained computer-vision models for agriculture, built financial data prototypes, and developed real-time vehicle-routing software that reduced fleet size by 20% and staff costs by 50% for the same order volume.
