At George Mason University, I developed an agentic AI pipeline to automate FDA violation analysis, reducing review time by 70%. I also created an NLP-based fraud detection system that achieved 90% precision on high-risk violation classification.
As a Data Engineer at Community Dreams Foundation, I designed data pipelines that consolidated more than five sources and reduced manual reporting effort by 60%. I also delivered real-time and batch pipelines for insurance data.
My projects include a PDF RAG chatbot that cut document review time by 40% and a real-time retail data pipeline. In my research at George Mason University, I built machine learning workflows for defect analysis, including an LSTM model that achieved 73% multi-class accuracy.

