This publish is cowritten by Renata Salvador Grande, Gabriel Bueno and Paulo Laurentys at Rede Mater Dei de Saúde.
The rising adoption of multi-agent AI methods is redefining vital operations in healthcare. In giant hospital networks, the place hundreds of selections instantly influence money circulate, service supply instances, and the danger of declare denials, the flexibility to observe, monitor, and govern AI brokers has change into important for operational sustainability. That is the journey of Rede Mater Dei de Saúde, which is implementing its suite of 12 AI brokers utilizing Amazon Bedrock AgentCore, a complete service that gives agent runtime, instrument integration, reminiscence administration, and built-in observability for manufacturing AI brokers.
About Rede Mater Dei de Saúde
With 45 years of historical past, Rede Mater Dei is one in every of Brazil’s most revered healthcare establishments, working amenities in Belo Horizonte, Betim-Contagem, Nova Lima, Salvador, Uberlândia, Goiânia, Feira de Santana, and a brand new mission underway in São Paulo. The group combines know-how, superior analytical intelligence, and high-complexity care to ship patient-centered outcomes and operational excellence.
Addressing a structural problem in Brazilian healthcare
In 2024, declare denials in Brazil reached alarming ranges, in accordance with the Nationwide Affiliation of Non-public Hospitals (Anahp): the sector common jumped from 11.89% to fifteen.89%, representing as much as R$ 10 billion in unreceived revenues. Like many establishments, Rede Mater Dei confronted operational challenges:
- Guide processes have been usually dealt with by tons of of operational employees.
- Fragmented processes have been characterised by unstructured and dispersed information.
- Excessive-turnover groups dealt with these procedures as a result of the repetitive nature of duties drove employees away.
- Advanced verifications that demanded intense, fixed consideration created inconsistencies and rework in weak levels of the method.
These weaknesses instantly impacted the income cycle, from credentialing to billing, and uncovered the group to the identical dangers pressuring the complete sector to extend denials. With help from A3Data and AWS, Rede Mater Dei launched a metamorphosis program to assist scale back the causes of denials, speed up analyses, and consolidate a ruled, observable, scalable, and high-quality operation via AI brokers.
Suite of 12 AI brokers deployed on Amazon Bedrock AgentCore Runtime
Rede Mater Dei, along with A3Data and the AWS Generative AI (GenAI) Innovation Middle, structured a program that includes a whole suite of 12 AI brokers designed to cowl the complete hospital income cycle. This suite created a “digital power” the place AI brokers understand, determine, and act as autonomously as doable, in an orchestrated, steady, and auditable method.
Among the many first applied brokers of the 12 deliberate are:
- Contracts Agent: Centralizes and constructions complicated contractual guidelines beforehand scattered throughout disparate paperwork.
- Parameterization Agent: Mechanically interprets guidelines into the hospital’s enterprise useful resource planning (ERP) system, serving to scale back human errors and speed up updates.
- Authorization Agent: Automates requests, validations, and interactions with well being insurers.
The brokers are executed on Amazon Bedrock AgentCore Runtime, which gives the safe, serverless internet hosting atmosphere for deploying, operating, and scaling AI brokers and instruments.
The group organized the structure into three complementary layers:
- DEL (Knowledge Execution Layer): Organizes information from a number of sources right into a structured information lake.
- AEL (Agent Execution Layer): Orchestrates and executes brokers in an built-in method.
- TCL (Belief and Compliance Layer): Applies governance, safety, and compliance alignment, facilitating traceability.
To control AI brokers, Rede Mater Dei partnered with A3Data and the AWS GenAI Innovation Middle. Collectively, they constructed the complete vital execution and governance layer for brokers on Amazon Bedrock AgentCore, which grew to become the operational coronary heart of the suite. This mission is a pioneering initiative in Latin America. It exams AgentCore Evaluations in a complete, large-scale AI resolution for a high-impact healthcare enterprise utility.
Why Amazon Bedrock AgentCore?
As a part of Amazon Bedrock, AgentCore is a complete set of providers offering the inspiration for agentic use instances. It provides instruments, deployment, and observability, together with the modular capabilities wanted to deploy, function, and refine AI brokers at scale and with safety. Elements embrace:
The preliminary implementation targeted on including a monitoring and enchancment layer to the present multi-agent AI resolution via AgentCore Observability and AgentCore Evaluations.
Utilizing AgentCore Evaluations, Rede Mater Dei added a monitoring and analysis layer to the answer. This layer helps steady enchancment of multi-agent AI methods with measurable, managed efficiency and excessive accuracy.
By this service, it’s doable to judge metrics and indicators thought of international greatest practices, corresponding to correctness, utility, precision, security, goal success charge, and context relevance.
This analysis construction gives measurement and traceability for AI brokers. These capabilities assist preserve stability, resilience, predictability, and regulatory adherence within the healthcare atmosphere.
Structure with AgentCore Evaluations
The answer was designed and applied as proven under, already incorporating AgentCore Evaluations:
Structure AgentCore Evaluations
517% ROI and structural effectivity throughout a number of processes
The preliminary part, anchored in AgentCore Observability and AgentCore Evaluations, enabled Rede Mater Dei to realize measurable positive aspects and lay the groundwork for a safer, extra predictable, and data-driven AI operation. The preliminary part delivered robust monetary and scientific outcomes: 517% return on funding (ROI) within the first 4 months, 66% discount in authorization time, and 33% discount in surgical procedure begin instances. Past these positive aspects, the AgentCore governance layer expanded the establishment’s capability to function and evolve its brokers with management and transparency.
Governance and compliance
Structured observability offered full traceability for vital income cycle choices, creating an immutable audit path for each interplay, rule utilized, and motion taken by brokers. This helped scale back regulatory dangers, strengthen operational safety, and simplify inside and exterior verifications—particularly in delicate processes like contracts, authorizations, and billing.
Operational effectivity
With unified telemetry, groups diminished the time spent figuring out and resolving failures, replicating impacts seen in different use instances with targets of as much as 50% discount in incident decision time. Groups can instantly spot anomalous behaviors, efficiency drops, or inconsistencies in brokers, serving to speed up steady enchancment cycles and enhance reliability in workflows instantly affecting denial danger.
Strategic decision-making
Actual-time visibility into key efficiency indicators (KPIs) covers the amount of automated analyses, projected monetary impacts, processing pace, estimated denial danger, profitable validation charges, and metrics per insurer. These insights remodeled operational information into quicker, extra exact govt choices. They information rule changes, backlog prioritization, group scaling, and surgical interventions in workflows most impacting income and effectivity. Collectively, these outcomes reveal that the mixture of Mater Dei’s agent suite and AgentCore helps not solely ship instant positive aspects but additionally helps set up the inspiration for a extra sturdy, auditable, and scalable hospital operation able to supporting the community’s enlargement and tackling Brazil’s structural problem of declare denials.
Past achieved outcomes, the mission has change into a worldwide reference, featured in a keynote speech by Ruba Borno (VP, AWS Specialists and Companions) at AWS re:Invent 2025 in Las Vegas, showcasing that remodeling the income cycle will not be solely doable: it’s measurable, fast, and able to producing substantial returns.
Testimonials
“Along with A3Data, we’re remodeling a historic business problem with a extra analytical, structured, and innovation-driven strategy. Our focus is to reinforce accuracy, predictability, and agility in vital income cycle levels, lowering variability and strengthening operational and monetary effectivity for the community. This improved consistency naturally interprets right into a smoother affected person expertise, pushed by extra organized and technically sturdy processes.”
– Renata Salvador Grande, Vice President of Industrial and Advertising and marketing, Rede Mater Dei de Saúde
Concerning the authors
Renata Salvador Grande
Renata Salvador Grande is the Vice President of Industrial and Advertising and marketing at Rede Mater Dei de Saúde. She is a lawyer with an MBA and Grasp’s in Advertising and marketing from Hult Worldwide Enterprise Faculty, and govt schooling from MIT. With almost 20 years in healthcare—together with roles at HCor and management positions at Rede Mater Dei throughout the income cycle—she chairs the A3Data Board, coordinates the Anahp’s Insurer Relations Group, and sits on the FIEMG strategic council.
Gabriel Bueno
Gabriel Bueno is the Lead Mission Guide and Companion at A3Data. With 17 years of expertise managing complicated initiatives, Gabriel is a Companion at A3Data, main mission consulting and superior resolution manufacturing. He has consulted for main corporations in healthcare, tourism, finance, and automotive sectors on Superior Analytics and Generative AI.
Paulo Laurentys
Paulo Laurentys is the Director of Operations and Companion at A3Data. He’s licensed by AWS in Generative AI (GenAI). With over 20 years in know-how and consulting, Paulo has led high-value initiatives for big enterprises utilizing rising applied sciences. He holds worldwide certifications from MIT, Johns Hopkins, Kellogg Northwestern, AWS, and EXIN Holland, and has held management roles at Inter and Accenture.
Lenilson Vilas Boas
Lenilson Vilas Boas is a Options Architect at AWS with a level in Laptop Science and specialization in Data Safety. Lenilson holds a Grasp’s in Synthetic Intelligence and instructing expertise. He guides AWS companions in growing and implementing cloud options tailor-made to shopper wants.
Evandro Franco
Evandro Franco is a Sr. Knowledge Scientist engaged on Amazon Internet Companies. He’s a part of the World GTM group that helps AWS prospects overcome enterprise challenges associated to AI/ML on prime of AWS, primarily on Amazon Bedrock AgentCore and Strands Brokers. He has greater than 18 years of expertise working with know-how, from software program growth, infrastructure, serverless, to machine studying. In his free time, Evandro enjoys enjoying along with his son, primarily constructing some humorous Lego bricks.
