This demo offers a hands-on exploration of how to streamline insurance claim processing by leveraging unstructured data, such as images, PDFs, and free text, using MongoDB Atlas, LLMs and embedding models. You’ll see live, interactive examples of vector search and Retrieval-Augmented Generation (RAG) applied to real-world challenges in claim management. The demo also includes architecture walkthroughs, with code available for attendees to inspect, replicate, and adapt to their own projects. Whether you're looking to improve operational efficiency or explore advanced AI applications in insurance, this session provides practical insights and a strong foundation for getting started.