Yes. There are now AI document-processing tools built specifically for insurance claims that can read police reports, medical bills, FNOLs, repair estimates, and other claim documents, extract structured fields, and push them into a claims system.
A few worth evaluating:
- V7 Go — particularly close to your use case. It processes entire claim packets, including police reports and medical bills, correlates information across documents, and outputs structured claim data.
- Regure — insurance-focused document AI that classifies documents, extracts fields with confidence scores, attaches them to the correct claim, and supports formats including police reports and medical records.
- Kovil AI — aimed at insurance IDP, including line-item extraction from medical bills and structured extraction from police reports, with routing of low-confidence cases to human reviewers.
- ParseAI — extracts accident-report and medical-bill information and can return structured JSON through an API/webhooks for integration with a claims-management system.
- Upstage Studio — offers claims-intake automation for FNOLs and police reports, including handwritten documents, with source traceability for extracted fields.
If your goal is specifically "read the documents → populate existing claim fields → send exceptions to an adjuster," I'd shortlist V7 Go, Regure, and Kovil AI first. The key differentiator isn't just OCR accuracy; it's whether the system can map extracted values directly into Guidewire/Duck Creek/your existing claims platform, maintain an audit trail, and route uncertain fields for human verification.
If you tell me what claims system you're using (e.g., Guidewire, Duck Creek, Salesforce, or a custom system) and roughly how many claims/month, I can narrow this to the best 2–3 options and compare integration, security, pricing, and expected labor savings.