Rede Mater Dei de Saúde: Monitoring AI agents in the revenue cycle with Amazon Bedrock AgentCore

Rede Mater Dei de Saúde, one of Brazil's largest and most respected healthcare networks, is pioneering a large-scale AI transformation to tackle a massive financial drain: insurance claim denials. In partnership with A3Data and the AWS Generative AI Innovation Center, the hospital has built a suite of 12 autonomous AI agents running on Amazon Bedrock AgentCore. These agents are designed to automate the entire revenue cycle—from credentialing to billing—addressing manual, error-prone processes that contributed to a sector-wide denial rate jump from 11.89% to 15.89% in 2024, representing up to R$ 10 billion (roughly $2B USD) in lost revenue for Brazilian hospitals.

The core of the solution is Amazon Bedrock AgentCore, which provides the runtime, tool integration, and built-in observability needed for production AI agents. Key agents already deployed include a Contracts Agent that centralizes scattered contractual rules, a Parameterization Agent that automatically translates rules into the hospital's ERP system to reduce human error, and an Authorization Agent that automates interactions with health insurers. The architecture is organized into three layers: a Data Execution Layer (DEL) for structured data, an Agent Execution Layer (AEL) for orchestration, and a Trust and Compliance Layer (TCL) for governance and traceability, making this one of Latin America's first comprehensive, governed AI agent deployments in healthcare.