AWS launches agentic AI platform for US healthcare organisations

Amazon Web Services (AWS) has launched an AI‑powered platform in the US that aims to cut administrative workloads in healthcare and improve access to patient services.

The platform, called Amazon Connect Health, uses agentic AI to automate tasks such as patient verification, appointment scheduling, clinical documentation and medical coding.

AWS said the technology is designed to free clinicians and administrative staff from time‑consuming processes so they can focus more directly on patient care.

The launch marks the first purpose‑built healthcare solution developed by AWS and is designed to integrate with the electronic health record (EHR) systems used by healthcare providers.

According to AWS, healthcare staff can spend up to 80 per cent of call time compiling patient data across multiple systems, including verifying patient identities, booking appointments and reviewing medical histories.

This can lead to delays in accessing care and places healthcare teams under additional pressure.

Amazon Connect Health is intended to address these issues by handling routine administrative interactions automatically.

AWS said the system can understand natural language requests from patients, confirm identities, check insurance information, review provider availability and schedule appointments in real time.

The platform combines the capabilities of Amazon Connect with direct connections to EHR systems, allowing it to retrieve and process patient data while escalating more complex requests to human staff when needed.

AWS added that the platform is designed to operate around the clock, enabling patients to book appointments without waiting on hold or leaving voicemail messages.

Beyond scheduling, AWS said the platform is intended to support clinicians throughout the entire care journey.

Before appointments, it can compile patient histories and highlight relevant information such as active conditions, recent medical events and longer‑term trends.

During consultations, and with patient permission, the system can transcribe conversations between doctors and patients and generate draft clinical notes in real time for review.

Following a visit, the platform can create patient‑friendly summaries and generate medical billing codes linked to supporting evidence from the consultation or medical records. According to AWS, this allows healthcare providers to prepare billing submissions within minutes rather than hours or days.

The company said healthcare organisations that have begun using the platform are already reporting efficiency improvements.

At UC San Diego Health, which manages around 3.2 million patient interactions each year, the technology has helped reduce call handling times by approximately one minute per interaction.

AWS estimates that the platform has diverted around 630 hours of staff time each week from patient verification tasks to direct patient support, while reducing call abandonment rates by 30 per cent.