AI Claims Assistant

Insurance claims can involve large amounts of policy and claim documentation that representatives need to understand quickly.

I built a RAG-based assistant that lets representatives ask questions about these documents and lets customers provide information naturally while the system helps complete claim documents.

Documents are processed, split into useful sections, embedded, and stored for semantic retrieval.

When a question comes in, relevant information is retrieved and provided to the LLM so the response is grounded in the available claim and policy documents.