At Kemper Insurance, I developed claims risk models that improved high-risk claim identification accuracy by 27%. I also built feature engineering pipelines for more than 20 million claims and policy records, reducing preparation time by 40%.
I designed a RAG solution using Azure OpenAI, GPT-4, LangChain, and Azure AI Search to retrieve insurance policies and claims documentation, reducing document search time by 65%. I also developed document intelligence solutions that reduced manual review effort by 55%.
At Beal Bank, I developed credit risk and loan default models to support lending decisions, improving predictive accuracy by 18%. I used model validation and monitoring practices to support stability and regulatory compliance.
At Mayo Clinic, I developed machine learning and NLP solutions for patient risk prediction and clinical analytics. At Macy's, I worked on customer personalization and recommendation systems, improving recommendation relevance by 15%.
