At Ernst & Young (EY), I deployed a production RAG knowledge assistant across 500K+ enterprise documents, improving information retrieval accuracy by 40%. I used LangChain, FAISS, Transformer LLMs, and SAP Databricks.
I also engineered a predictive demand-forecasting pipeline that raised F1-score from 0.61 to 0.82, and delivered SAP Analytics Cloud dashboards that reduced ad-hoc reporting requests by 30%. I work across machine learning, Generative AI, and enterprise data platforms.

