At Northern Trust, I build quantitative credit risk, fraud detection, and portfolio analytics models using transactional and market data. My work has improved credit risk prediction accuracy by 24% and reduced transaction false positives by 21%.
I architect PySpark and Hive ETL pipelines processing more than 60 million financial records daily, reducing data latency by 37%. I also develop RAG pipelines, vector database integrations, and fine-tuned transformer models for financial reports, filings, investment research, and trading signals.
Previously at Dell Technologies, I delivered demand forecasting, customer segmentation, pricing optimization, and BI solutions that improved forecast accuracy, increased campaign effectiveness, and generated a $3.2M quarterly revenue increase.
