Honors Program
Honors Program: Embargoed Theses
Date of this Version
5-2025
Document Type
Thesis
Citation
Daup, S., Jain, A., Siner, M., Tarkian, A., & Walters, C. 2025. Claims genAI Copilot. Undergraduate Honors Thesis. University of Nebraska-Lincoln.
Abstract
The team was tasked with enhancing the efficiency of Mutual of Omaha’s claims processing operations by introducing a generative AI-driven assistant. Historically, claims benefit specialists relied on legacy systems and manual review of dense policy and rider documentation—methods that are not only time-consuming but also difficult to scale. In collaboration with Mutual of Omaha’s internal GenAI team, the Design Studio team developed an AI-powered solution designed to provide fast, accurate access to policy information through natural language queries.
The team implemented a Retrieval-Augmented Generation (RAG) architecture that leverages large language models via AWS Bedrock to retrieve and deliver form-specific content in real time. By integrating document-level metadata filtering, semantic embedding, and a user-friendly Gradio interface, the solution enables claims associates to interact with hundreds of complex insurance documents in a conversational and guided manner. This project sought to modernize the existing system with intelligent automation that reduces claim processing time, enhances training for new hires, and improves the overall user experience for employees engaging with critical policy content.
Comments
Copyright Seth Daup, Aditya Jain, Mia Siner, Amir Tarkian, and Cole Walters 2025.