AWS has launched Amazon Connect Health, an agentic AI solution targeting the administrative bottlenecks plaguing healthcare providers.
This marks the first purpose-built AWS solution for the healthcare industry
Healthcare contact centers are among the most operationally strained in any sector.
AWS’s own research suggests that health system staff spend up to 80% of call time on manual data compilation, toggling between fragmented tools to verify identities, check insurance eligibility, and piece together patient histories.
According to AWS, 89% of patients cite care navigation challenges, such as difficulty scheduling, long wait times, and access barriers, as reasons for switching providers.
In a sector already stretched thin, this becomes both a churn problem and a care problem.
Amazon Connect Comes to Healthcare
Amazon Connect Health integrates directly with Electronic Health Records (EHRs) and layers agentic AI across the patient contact journey, with capabilities including real-time patient verification, 24/7 appointment scheduling via natural language voice interaction, and live insurance eligibility checks.
When a patient calls to book an appointment, the system can understand the request, confirm their identity, check coverage, and lock in the booking in real time – all without a human agent ever needing to pick up.
Where things get more complex, the solution escalates to a human, with health systems able to configure their own handoff protocols.
The combination of agentic automation and human oversight isn’t a new concept for contact center leaders, but seeing it deployed at this scale, in this setting, with actual numbers attached, is another thing entirely.
The move into healthcare suggests AWS sees significant runway in applying its established CX infrastructure to sectors where fragmented data and compliance complexity have historically made the contact center experience particularly painful.
The Results That Make the Case
The early numbers from UC San Diego Health are hard to argue with.
The health system manages 3.2 million patient interactions annually across 45 contact centers.
After deploying Amazon Connect Health, it reported saving one minute per call, diverting 630 hours weekly away from patient verification to direct patient assistance, and cutting call abandonment rates by 30%, with some departments seeing reductions as high as 60%.
While the abandonment rate reduction is undoubtedly impressive, arguably the more interesting framing in the official blog is around what the time savings actually unlock.
With staff no longer tied up in verification, they are able to redirect their attention to interactions that genuinely require human judgment.
It’s an argument contact center leaders have been making for years; UC San Diego Health is now a real-world example to back it up.
Amazon One Medical, a national primary care organization and Amazon subsidiary, has also been putting the solution to work in live clinical settings.
One feature worth flagging for the contact center audience is the pre-visit patient insights capability, which surfaces relevant medical history and context before an appointment.
When clinicians walk in prepared, the need for follow-up calls, clarifications, and repeat contacts through the contact center drops.
Amazon One Medical has expanded its use of the solution this year to include intelligent medical coding, further reducing the post-visit admin that typically generates downstream patient queries.
Elsewhere, Netsmart – a major EHR provider for community-based care organizations – has seen ambient documentation adoption increase by 275% since deploying the solution, with providers spending less time on admin, more time with patients, and improved staff retention as a by-product.
What This Means Beyond Healthcare
On the trust and compliance side, AWS offers over 130 HIPAA-eligible services and employs what it describes as “evidence mapping,” which links every piece of AI-generated output back to its source, whether that’s a conversation transcript or a billing guideline.
The AI is trained on healthcare-specific data rather than general-purpose models, with multi-step safety and accuracy evaluations built in.
For CX and contact center professionals, there’s a broader story here beyond the healthcare specifics.
Amazon Connect Health is essentially AWS taking a platform it has spent years refining for general contact center use and stress-testing it in one of the most demanding service environments around.
If the early results hold, it is not difficult to imagine financial and insurance services – sectors with their own long history of compliance-driven contact center friction – being next in line.
