AWS announces Amazon Connect Health

SEATTLE, Wash. – AWS is diving more deeply into the healthcare sector with the announcement of Amazon Connect Health, which it’s calling an agentic AI solution that handles high-volume administrative tasks such as appointment scheduling, clinical documentation, and medical coding. Currently it’s only available in the U.S.

“Rather than simply streamlining existing processes to work faster, Amazon Connect Health enables entirely new ways of working,” said Naji Shafi (pictured), general manager and director of Healthcare AI at AWS.

“Clinicians focus on care while insights surface seamlessly, patients receive scheduling assistance instantly without staff manually compiling data, and medical billing completes in minutes instead of days without sacrificing accuracy.”

Amazon Connect Health builds on Amazon Connect, AWS’s AI-powered solution for delivering personalized customer experiences at scale that serves tens of thousands of businesses worldwide and handles over 16 million interactions daily.

For healthcare provider organizations
Amazon Connect Health offers agentic AI capabilities for patient engagement that are pre-integrated with Amazon Connect. Available today:

• Patient verification (generally available): Conversational patient identity verification with real-time EHR integration to eliminate manual record lookup.
• Appointment management (preview): Natural language voice scheduling with 24/7 availability, real-time insurance verification, and EHR integration. Patients are able to book, reschedule or cancel without hold times.

Key benefits:

• 24/7 patient access with minimal wait times.
• Reduced administrative burden by streamlining scheduling and verification, freeing staff to focus on patient care.
• Robust reporting, analytics, and multi-channel engagement because it is built natively into Amazon Connect with world-class reporting, analytics, and multi-channel engagement, rather than a siloed solution.
• Purpose-built for healthcare with understanding of call context, patient preferences, and seamless escalation to staff when situations require a human touch.

For healthcare tech builders
Builders in EHR companies, healthcare ISVs (Independent Software Vendors), and tech-enabled providers can leverage a unified SDK to directly integrate Amazon Connect Health’s point of care capabilities into their existing workflows. Each capability is fully managed and modular, enabling builders to incrementally add agentic capabilities through the same SDK without requiring new integration projects.

Available today:

• Patient insights (preview): Surfaces visit-specific insights including patient summaries, health events since last visit, and HCC (Hierarchical Condition Categories) recapture from structured and unstructured longitudinal patient records.
• Ambient documentation (generally available): Generates clinical notes from patient-clinician conversations in real time, automatically formatted into existing EHR templates. Supports 22+ specialties and offers traceability from generated notes to transcripts.
• Medical coding (preview): Generates ICD-10 and CPT codes from clinical notes with confidence scores and full source traceability.

Key benefits:

• Reduced administrative burden for clinicians with output linked back to the source data to drive trust. Clinicians get more face time with patients.
• Accelerated revenue cycles for provider organizations with medical codes ready to review by the time each patient visit ends.
• Speed development timelines and reduce undifferentiated heavy lifting for engineering teams, who no longer need to spend months or years building these capabilities, optimizing AI accuracy, or maintaining them as the underlying AI capabilities evolve.
• Integrate AI into existing workflows in less than a week. Unlike solutions requiring screen and workflow changes, the unified SDK enables direct integration into EHR screens and digital front doors.

According to AWS, Amazon Connect Health builds trust and safety into the product. Every patient insight, clinical note, and billing code traces back to its source transcript or patient chart data.

Clinicians tap any AI output and view the underlying evidence immediately. The solution combines supervised fine-tuning, reinforcement learning, LLM reasoning, and information retrieval trained with healthcare domain expert guidelines and real-world data.

Critical capabilities undergo rigorous validation through manual evaluation and automated testing, meeting AWS standards for robustness, safety, and scalability. Patient-facing verification and scheduling agents automatically escalate to a human when needed.

Making human-AI collaboration real for healthcare
Agentic AI in healthcare is only as good as the data it can access. However, healthcare data often arrives fragmented across multiple systems and formats. For customers who do not have their data in a unified format, Amazon Connect Health addresses this with pre-built connectivity into AWS HealthLake, a petabyte-scale healthcare data layer that transforms disparate records into actionable intelligence.

Amazon Connect Health capabilities, such as patient insights, integrate with AWS HealthLake in a few clicks. AWS HealthLake also reduces manual effort for normalizing external records (e.g., transforming CCDA files into FHIR format) through a purpose-built data transformation agent, which just launched in preview.

Additionally, the company is working with partners such as Redox who have pre-existing integration with 100+ EHRs and 35+ HIEs (Health Information Exchanges) to help customers seamlessly integrate their data. As a combination of these capabilities, customers can point agents to unified data sources to receive context-driven intelligence, enabling clinicians to make informed decisions without piecing together information from multiple systems.