What challenges did viesure Innovation Center face before AI implementation?

Reduce the time insurance claims handlers spend reviewing and understanding fragmented claims data and documents. Ensure complete GDPR compliance, data sovereignty, and data security when using powerful cloud-based AI models on sensitive personal data. Build user trust with phased rollout, human-in-the-loop approach, and alignment with EU AI Act requirements for AI system governance.

What was the impact of the AI solution?

Achieved 96% accuracy in automated identity document extraction. Significantly reduced insurance claims review time. Ensured strong data security, privacy, and GDPR compliance, enhancing user trust and acceptance of AI-assisted workflows.

Source

GCP

viesure Innovation Center: GDPR-Compliant AI for Claims Automation at Wiener Städtische Versicherung with Vertex AI and Gemini

Use case typeIntelligent document processingUpdated Jun 13, 2026

viesure Innovation Center, the innovation hub of Wiener Städtische Versicherung AG in Austria, developed a GDPR-compliant AI system to reduce insurance claims review time while ensuring data sovereignty and regulatory compliance. The Smart Claim Assistant uses Google Cloud Vertex AI and Gemini to analyze and summarize claim documents, improving claims handlers' efficiency without automating final decisions to maintain trust. A multi-layered security architecture with BYOK encryption, no data persistence in LLM processing, and ephemeral GPU computation ensures data protection and compliance with EU AI Act governance. The solution also includes a hybrid identity document extraction system combining Gemini, Cloud Vision API, and Document AI achieving 96% accuracy.

Organization: viesure Innovation Center
Industry: Insurance
Location: Austria

Reported outcomes

96%

accuracyQuality & accuracy

Strategic outcomes

Customer experience & trustImproved trust in AI-assisted workflowsRisk & complianceAchieved GDPR-compliant claims automationNew product / capabilityBuilt AI claims summarization assistantNew product / capabilityCreated hybrid identity document extraction

Catalog median for quality & accuracy deployments: +90% across 282 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 3

1AI-assisted claims document analysis
2GDPR-compliant AI processing
3Secure identity document extraction

Reduce the time insurance claims handlers spend reviewing and understanding fragmented claims data and documents.
Ensure complete GDPR compliance, data sovereignty, and data security when using powerful cloud-based AI models on sensitive personal data.
Build user trust with phased rollout, human-in-the-loop approach, and alignment with EU AI Act requirements for AI system governance.

Developed Smart Claim Assistant leveraging Vertex AI and Gemini to analyze claim documents and provide summarized structured dashboards to claims handlers.
Implemented a hybrid model for identity document extraction combining Gemini, Cloud Vision OCR, and Document AI technologies.
Employed a multi-layered security architecture with customer-managed encryption keys (BYOK), temporary GPU computation without data persistence, and regional data processing in Europe.
Established comprehensive AI governance including risk-based assessments, employee training, human-in-the-loop workflows, and transparent documentation to meet EU AI Act standards.

Technologies

Vertex AI Gemini Cloud Vision API Document AI BigQuery Cloud Storage Cloud Key Management

Achieved 96% accuracy in automated identity document extraction.
Significantly reduced insurance claims review time.
Ensured strong data security, privacy, and GDPR compliance, enhancing user trust and acceptance of AI-assisted workflows.

Architecture

The technical architecture integrates Google Cloud Vertex AI, Gemini, Cloud Vision API, Document AI, BigQuery, Cloud Storage, and Cloud Key Management to securely process claim documents with ephemeral GPU computation and BYOK encryption.

Sources & evidence1

Groundedness: 5/5

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence