When customers with complex, multi-touch claims called Nationwide's contact centers, service representatives faced a structural bottleneck: before answering a single question, CSRs had to manually read through every log note from every prior interaction — a process taking three to four minutes per call. For a Fortune 100 insurer handling high volumes of auto, home, and life claims, that wait translated directly into customer frustration and reduced capacity for empathetic service. Traditional OCR technology compounded the issue by slowing document ingestion and requiring manual field mapping, adding friction to an already time-intensive workflow. The cumulative effect was that CSRs spent their most critical customer-facing minutes catching up on history rather than resolving problems.
Nationwide deployed a Generative AI-powered Claims Log Notes Summarization engine that synthesizes the full interaction history of any claim into an instant, structured summary surfaced the moment a CSR pulls up a customer record. Built on top of Nationwide's existing Guidewire ClaimCenter platform and their newly migrated Genesys Cloud contact center infrastructure — which provided enhanced NLP and voice recognition — the tool required no complex model training, leveraging generative AI's native summarization capability. A complementary Call Transcript Summary tool was introduced shortly after to auto-populate after-call work fields from call transcripts, further reducing post-interaction handle time. Both tools were designed with a human-in-the-loop architecture, positioning AI as an assistant to CSR judgment rather than a replacement for it.
Claims Log Notes Summarization has been in continuous production for over a year and achieved strong adoption among CSRs, who report measurable improvements in speed-to-answer. Customers experience shorter hold times before receiving accurate claim status updates. Key outcomes include:
Nationwide's CTO attributes the Keynova ranking in part to these AI-driven investments in the claims experience.
Have a similar implementation?
Share your customer's AI results and link it to your vendor profile.
Submit a case study →