AWS Launches Amazon Connect Health AI Platform to Support Healthcare Administration

Health AI;Amazon’s cloud computing division, Amazon Web Services (AWS), announced the launch of an artificial intelligence-enabled platform designed to assist healthcare providers with administrative tasks and improve patient access to care. The platform, called Amazon Connect Health, introduces agentic AI capabilities intended to manage processes such as appointment scheduling, patient verification, and medical documentation.

According to AWS, the system integrates with electronic health record systems used by clinicians. Through this connection, the platform can assist with verifying patients, scheduling appointments, compiling medical histories, documenting clinical encounters, and generating medical codes used for billing. The company said the platform operates around the clock, allowing appointments to be booked instantly and escalating complex situations to staff when required.

The solution uses natural language voice technology to interpret patient requests during calls. It analyzes the purpose of the call, confirms patient identity, checks insurance coverage, and reviews patient and provider availability before scheduling an appointment while the patient remains on the line. The platform combines Amazon Connect, the company’s AI-powered customer experience solution, with real-time connections to electronic health record systems.

AWS said the platform can be deployed in days because of native integrations with electronic health record systems and Amazon Connect contact center tools. It also includes prebuilt connectivity to more than 100 electronic health record systems through data integrator partners and provides a fully managed unified software development kit intended to simplify implementation for healthcare providers and health technology companies.

The company said the platform was designed to address administrative workloads that can take up a large portion of healthcare staff time. Colleen Aubrey, senior vice president of applied AI solutions at AWS, described the situation in a blog post. She wrote, “In conversations with large health systems, AWS has found that staff spend up to 80% of call time on manual data compilation across fragmented tools.”

Beyond scheduling functions, the system can review a patient’s medical history before a doctor visit and provide clinicians with a concise summary. According to AWS, these summaries may include information on active conditions, recent events, trends over time, and chronic conditions that could be relevant for care gap closure and billing accuracy.

The platform also expands AWS’s ambient documentation technology. It can transcribe conversations between doctors and patients during medical visits and generate draft clinical notes in real time for provider review. After the appointment, the system can produce patient-friendly summaries and suggest medical billing codes, with each code linked to the supporting evidence used to generate it.

AWS said the platform includes a feature called evidence mapping designed to provide transparency in AI-generated outputs. This function links every output generated by the system to its exact source, which may include call transcripts, medical records, or billing guidelines. According to the company, clinicians reviewing summaries or documentation can access the underlying source material in order to verify the information before completing records.

The system relies on specialized learning approaches that apply supervised fine-tuning and reinforcement learning techniques to healthcare-specific datasets and guidelines. AWS said model performance undergoes multiple evaluation steps designed to measure safety and accuracy. These assessments include automated evaluations in which one AI system checks another system’s output as well as clinician-in-the-loop validation processes.

Some health systems have already deployed capabilities from the platform. UC San Diego Health reported that after implementing the technology, it saves approximately one minute per call and has reduced call abandonment rates by 30%, reaching as high as 60% in some departments. The organization manages about 3.2 million patient interactions each year and has diverted 630 hours weekly from patient verification tasks to direct patient assistance.

AWS also said that Amazon One Medical has used the platform’s documentation feature in more than one million visits, with strong clinician adoption and regular weekly usage. The company said the organization plans to expand its use of the system to intelligent medical coding this year.

Another healthcare technology company, Netsmart, has integrated Amazon Connect Health with its systems. According to AWS, Netsmart reported that adoption of ambient documentation capabilities increased by 275% among its users.

AWS said the platform was built with security and compliance considerations. Aubrey noted that the company maintains 130 services that are eligible under the Health Insurance Portability and Accountability Act and holds certifications for global IT and compliance standards.

How Health AI Improves Healthcare Administration

The Health AI tools within Amazon Connect provide intelligent automation that can assist healthcare organizations with high-volume communication tasks. With Health AI, healthcare staff can handle patient calls more efficiently while reducing manual processes.

Using machine learning and natural language processing, Health AI can analyze patient requests, route calls to the right department, and provide automated responses for common healthcare questions. This allows medical staff to focus more on patient care while Health AI handles repetitive administrative duties.