In highly regulated environments such as the healthcare and insurance sectors, providers often need to perform tasks like analyzing health risks and verifying customer claims, while keeping data strictly within the relevant jurisdiction.
The solution described in this document provides a blueprint for solving these challenges with Cloud de Confiance by S3NS. Insurance providers can deploy BigQuery ML on Cloud de Confiance by S3NS to analyze health risk scores, model plans, and claims databases in real time. Simultaneously, a local, open-weight Gemma LLM system performs verification on claim documentation stored in Cloud Storage. This approach ensures context-aware compliance and full auditability, without ever making external API calls or moving data outside the sovereign boundary.
You can deploy this solution by following the accompanying reference implementation with Terraform.
Target audience
This solution is designed for national healthcare agencies or regulated insurance providers. It serves the following stakeholders:
- Auditors and claims investigators who review showroom dashboards to verify submitted claim documents, audit flagged cases, and interact with the AI assistant.
- Data scientists who securely access raw data using JupyterLab environments to build, validate, and run health risk analysis models.
Core capabilities
- Sovereign health modeling: Use BigQuery ML to screen claims and customer datasets to identify health risk factors without moving data outside the sovereign boundary.
- AI claim verification: Run localized document checks using open-weight Gemma LLM to verify claim details against database records.
- Interactive chatbot widget: Query active database claims and receive real-time, context-aware answers using the AI assistant chatbot.
- Seamless JupyterLab analytics: Access a dedicated Jupyter environment preloaded with direct database connection variables and Workload Identity bindings.
- Internationalization: Instantly switch UI language between English, French, and German.
Architecture
Components
The following technologies and Cloud de Confiance by S3NS services are used in this solution:
| Component | Tech | Purpose |
|---|---|---|
| Database | Cloud SQL | Storage for insurance claims, plans, and customers. |
| Other storage | Cloud Storage | Storage for claims documentation. |
| DWH | BigQuery | Data warehouse analytical tables and big data storage. |
| Model | BigQuery ML | Machine learning analysis on health risk scores. |
| LLM | Gemma | Open-weight model (google/gemma-3-27b-it) for claim document verification and chatbot widget. |
| Container infrastructure | GKE | Container infrastructure for the claim application and JupyterHub server. |
| Analysis | JupyterLab | Dedicated notebooks preloaded with active data. |
| App | Node.js | Web app showroom dashboard. |
Reference implementation
A reference implementation of this solution with Terraform is provided in GitHub. Note that this is a proof-of-concept prototype built for demonstration purposes, and the implementation is not audited or secured for production use cases.
For prerequisites and deployment instructions, see Sovereign Health Insurance Risk Analysis with BigQuery ML & Gemma.