ERIC - Agentic Underwriter for Health Insurance

ERIC helps health insurance companies implement, integrate, and operationalize Agentic AI-powered underwriting workflows on AWS.

This professional services engagement supports the deployment, configuration, integration, and production enablement of ERIC, an Agentic Health Underwriter built on LUMIQ’s LiteCone platform.

Based on the proposed solution architecture, ERIC uses AWS services including Amazon Bedrock, Amazon Textract, Amazon Elastic Container Service (Amazon ECS), Elastic Load Balancing, Amazon Simple Storage Service (Amazon S3), Amazon DynamoDB, Amazon Elastic Container Registry (Amazon ECR), AWS Secrets Manager, Amazon CloudWatch, AWS Identity and Access Management (AWS IAM), and Amazon Virtual Private Cloud (Amazon VPC).

ERIC is designed to provide end-to-end support for health insurance underwriting by combining intelligent document handling, verification, risk assessment, workflow orchestration, automated decisioning, and human-in-the-loop review.

The solution supports the processing of proposal forms, KYC documents, financial documents, medical questionnaires, medical reports, previous policy records, product configurations, underwriting guidelines, and external validation data.

ERIC supports the following underwriting activities:

Document and quality assessment, Document classification and tagging, Data extraction and structuring, Data validation and verification, Cross-document validation, Identification and correction of document issues, Pending requirement identification and handling, Tele-medical call analysis, KYC and external data verification, Credit bureau and third-party data integration, Medical risk assessment, Lifestyle and occupational risk assessment, Financial and moral hazard assessment, Fraud and document fraud detection, Risk correlation across applicant information, Automated underwriting decision support, Human-in-the-loop review and monitoring, Integration with downstream systems of record

ERIC helps underwriting teams identify relevant risk signals from large volumes of documents and information. It supports insurers in reducing manual review effort, improving consistency, responding faster to underwriting guideline changes, and managing seasonal or peak underwriting volumes. The proposed architecture deploys LiteCone application services using Amazon ECS within an Amazon VPC and private subnet. Elastic Load Balancing routes incoming requests to the application services.

Amazon Bedrock supports the AI capabilities used for underwriting analysis and decision support. Amazon Textract supports the extraction of text and structured information from proposal forms, medical records, financial documents, KYC documents, and supporting records.

Amazon S3 supports document and data storage, while Amazon DynamoDB supports application persistence and workflow-related data. Amazon ECR manages container images used by the solution. AWS Secrets Manager manages credentials and application secrets, AWS IAM controls access and permissions, and Amazon CloudWatch supports application monitoring, logging, and operational visibility.

The solution can integrate with policy administration systems, underwriting applications, document management systems, KYC services, credit bureaus, third-party data sources, tele-medical platforms, internal workflows, and downstream systems through inbound and outbound APIs and event-based integrations.