What challenges did the organization face before AI implementation?

Manual claim processing required agents to perform repetitive, time-consuming review of each request. Claims arrived as varied unstructured bills, invoices, pharmacy and clinic receipts, diagnoses, and supporting documents that required detailed examination and categorization. Existing technology could not support swift end-to-end automation, reducing customer service efficiency.

What was the impact of the AI solution?

70% of documents fed into the system are correctly extracted and interpreted. Claim documents are processed in near real time. Agents have more time for personalized advice and customer interactions.

Source

MicrosoftProductionEvidence: Medium60/100

Nordic Insurer Streamlines Claims Processing and Boosts Customer Service

Use case typeClaims automationUpdated Jun 13, 2026

An unnamed Nordic insurer engaged EY to modernize manual claims processing across large volumes of varied, unstructured documents that existing technology could not process swiftly through end-to-end automation. EY Fabric Document Intelligence, built on machine learning and Python and hosted in an EY-secured cloud, cleans scanned files and uses OCR and NLP to convert and classify unstructured data before transferring structured output to the core claims system. The implemented solution provides near real-time claims document processing, with 70% of documents fed into the system correctly extracted and interpreted, giving agents more time for personalized customer interactions.

Industry: Insurance
Published: June 2025

Planned next steps

The source says the organization aims to achieve: Supported global expansion goals.
Operational efficiency and customer service improvements support the insurer's global expansion goals.

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Documents correctly extracted and interpreted: 70%

EYJun 11, 2025Case studyInferred claimMedium evidence strength

Since the solution's implementation, a remarkable 70% of the documents fed into the system are correctly extracted and interpreted.

Last evidence check: Aug 9, 2026

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Not established
Provider: Microsoft
Maturity: Production
Linked source: EY

The insurance firm now benefits from near real-time processing of claim documents. Since the solution's implementation, a remarkable 70% of the documents fed into the system are correctly extracted and interpreted. The operational efficiency and customer service enhancements resulting from the implementation are sparking curiosity in other areas of the organization.

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 2 of 2

1Automated insurance claims processing
2Claims document digitization and structured-data integration

EY analyzed the insurer's technology infrastructure and business needs and tailored EY Fabric Document Intelligence to its requirements.
Built on machine learning and Python and hosted in an EY-secured cloud, the product cleans scanned images and performs preprocessing, document analysis, and layout analysis.
Optical character recognition (OCR) and natural language processing (NLP) convert and classify unstructured data before structured output is transferred to the core claims system.
Controlled confidence levels give the insurer transparency and control over the automation process.

Technologies

EY Fabric Document Intelligence Machine learning Python Optical character recognition (OCR)Natural language processing (NLP)EY-secured cloud