Nationwide

Commercial agribusiness underwriting at Nationwide required evaluating complex, multi-line policies spanning numerous vehicles, buildings, and locations across multiple states — all documented in dense, unstructured submissions. Underwriters spent significant time manually sifting through public company information and policy documentation before they could assess risk or respond to agents. Gaps in submissions often went undetected until late in the workflow, creating reactive bottlenecks and delayed agent responsiveness. This manual burden constrained underwriting capacity, limiting how many submissions the team could process and slowing business growth across the agribusiness vertical.

Nationwide deployed a suite of NLP-powered AI tools purpose-built for commercial agribusiness underwriters. The tools use natural language processing to automatically identify gaps in incoming submissions, surface summaries of public company information, and retrieve context from prior agent conversations — enabling proactive outreach before issues escalate. Rather than pursuing a monolithic deployment, the team took a minimal viable product approach: building and testing small, targeted tools incrementally with input from underwriting volunteers and subject-matter experts. Nationwide also rolled out Glean for internal knowledge retrieval, combining structured and unstructured information from internal repositories and external sources. Brad Liggett, president of the agribusiness vertical, designated a senior underwriter as an AI champion to guide peer adoption and translate technical capabilities into practical daily workflows. C-suite sponsorship, including from P&C group president Mark Berven, ensured alignment from leadership through execution.

The AI initiative reduced underwriting cycle time and improved Nationwide's responsiveness to agents by surfacing submission gaps and prior conversation context earlier in the workflow. Underwriters retained full decision-making authority — AI functioned as an aide, not a replacement. Key outcomes include:

• Faster submission reviews through automated gap detection and public information summarization
• Improved agent responsiveness via proactive outreach enabled by prior-conversation context retrieval
• Increased underwriting capacity without adding headcount, supporting business growth across the agribusiness vertical
• High adoption rates attributed to the peer-led AI champion model and incremental tool rollout

• Embedding a domain expert as AI champion — rather than relying solely on technologists — accelerates adoption by translating capabilities into peer-trusted, practical workflows.
• A minimal viable product approach builds trust incrementally and reduces deployment risk compared to large-scale, all-at-once rollouts.
• C-suite sponsorship paired with front-line involvement is essential; top-down mandate without underwriter participation stalls adoption.
• AI in underwriting delivers the most value as an augmentation layer — handling information retrieval, gap detection, and summarization — while preserving human judgment for final risk decisions.
• Winning hearts and minds is as critical as the technology itself; change management should be treated as a first-class workstream.