The Travelers Companies, Inc. announced this week that engineers and data scientists have developed TravelersLLM, a proprietary large language model tailored to its property/casualty business. Trained on millions of company documents, TravelersLLM “amplifies Travelers’ leading domain expertise by, among other things, enhancing underwriting analysis, accelerating research and model development, facilitating access to decades of institutional knowledge and improving workflows,” according to a company statement. The development reflects the company’s long-standing strategy of deploying advanced technology with proprietary data. TravelersLLM serves as a foundational capability for agentic applications across the enterprise and outperformed commercially available AI models in tests seeking answers to “tens of thousands” of insurance-related questions, delivering “higher-quality results at lower cost and with greater speed.” Mojgan Lefebvre, Executive Vice President and Chief Technology & Operations Officer, stated that TravelersLLM combines “vast amounts of well-curated data and our industry expertise with leading AI capabilities.” In a LinkedIn message, Lefebvre wrote that TravelersLLM works alongside frontier models to bring greater precision, context, and efficiency to use cases including underwriting, claims, and service, combining frontier AI capabilities with “decades of institutional knowledge, millions of proprietary documents, deep insurance expertise, and world-class engineering.”