Zelin Z.

Data Scientist @Mutual of Omaha

Seattle, WA, US

Developed a machine learning based scoring system that assessed claim complexity and boosted claim assignment efficiency by 70%, built a feature engineering and modeling pipeline with LGMB model, designed and implemented the cloud architecture for the application on AWS, developed Tableau dashboard to deliver model results for end users.• Built an agent-based GenAI chatbot to gather claim information and answer claim-related questions which improved claim processing efficiency by 50%, conducted prompt engineering to optimize LLM response quality which achieved 98% accuracy, designed and implemented the Function Calling in Action Group for the Bedrock Agent, created and configured KnowledgeBase and Guardrail via AWS CDK, assisted Gradio application testing and deployment.• Leveraged Generalized Linear Model to identify and predict customer groups with high lapse rate, wrote recursive SQL query to find the worst branch on broker hierarchy, provided actionable insights for business decision-making.

Beijing Normal University

Bachelor of Management

University of California, Berkeley, Haas School of Business

Exchange Student

University of California, Berkeley

Master of Engineering - MEng