Zero Touch Claims Optimizing P&C Insurance Claims Using AWS AI/ML
Property & Casualty (P&C) insurance companies including FWD Group and nib Group are reducing manual efforts in claims processing by adopting AI/ML technologies.
- Organization
- FWD Group
- Industry
- Insurance
- Location
- United States
- Published
- June 2022
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- FWD Group, nib Group
- Provider
- AWS
- Maturity
- Production
- Linked source
- AWS Industries Blog
High operational costs and low automation in claims workflows due to manual intensive processes
Primary read
Use case focus
Showing 3 of 4
- 1Claims Automation
- 2Predictive Analytics
- 3Computer Vision
- Implemented AI/ML-driven zero-touch claims workflow automation using multiple AWS services: Amazon Textract for document processing, Amazon SageMaker for predictive analytics, Amazon Rekognition for image and video analysis, Amazon Comprehend for natural language processing, Amazon Lex for conversational AI, AWS Step Functions for orchestration, and Amazon Connect for omnichannel customer engagement.
- Automated the entire claims lifecycle from first notice of loss to settlement, enhancing accuracy, speed, and customer experience.
Architecture
Covers an end-to-end claims processing architecture that leverages AWS AI/ML services through each phase including first notice of loss, image and video capture, loss assessment with computer vision, document processing, fraud detection with ML models, settlement automation, and multichannel customer engagement.
Sources & evidence1
- Customer explicitly identified
- Deployment status explicitly supported
- Technical implementation details available
The same organization appears in newer AI deployment evidence.
- Same organization re-documented as recently as 2026.
Measures whether this deployment's public evidence persists — not whether the system is still in production.
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?
Community
Comments
No published comments yet.