Amazon Web Services (AWS) has introduced a new artificial intelligence platform targeting the healthcare sector, designed to automate a range of administrative and clinical support tasks. The service, named Amazon Connect Health, was launched on Thursday, March 5, 2026, with the goal of alleviating the paperwork and phone traffic that often burdens clinics and hospitals.
The platform connects directly to electronic health record (EHR) systems and is built with what AWS terms "agentic AI"—software capable of autonomous action beyond simply responding to queries. Its initial features include patient verification and ambient clinical documentation, which are now generally available. Additional capabilities such as appointment management, patient insights, and medical coding are currently in a preview phase. The medical coding tool is specifically designed to generate ICD-10 and CPT billing codes for U.S.-based healthcare organizations.
In a blog post, AWS Senior Vice President Colleen Aubrey highlighted the significant time drain caused by current processes, noting that staff can spend up to 80% of their call time managing data across disparate systems. "Health care should be about people, but too often, the patient is overshadowed by the process," Aubrey wrote. Early customer results appear promising. UC San Diego Health reported saving approximately one minute per call using the system and has observed a decrease in abandoned calls. Amazon's own primary care service, Amazon One Medical, has utilized the documentation feature for more than one million patient visits.
Naji Shafi, head of healthcare AI at AWS, explained to Healthcare Dive that the platform targets high-volume, repetitive workloads that contribute to staff burnout. "Our healthcare workers are overburdened, drowning in administrative complexity," Shafi stated. The system includes safeguards, such as human review for documentation and codes, and employs a "large language model as a judge" where a secondary AI model assesses the output of the primary one. It can also detect patient frustration during interactions and escalate the call to a human agent.
AWS has set a pricing model starting at $99 per user monthly for the ambient documentation feature, covering up to 600 patient encounters. Additional usage incurs extra fees. Patient verification is priced at $0.15 per "action," defined as one complete conversation. The company is offering time-limited free trials for first-time users.
The launch intensifies competition in the healthcare AI and cloud workflow arena. On the same day, CVS Health announced a partnership with Alphabet's Google Cloud to develop Health100, another AI-driven platform scheduled for a 2026 launch, which will also utilize agentic AI technology. This move by AWS represents a strategic effort to further integrate Amazon's broader healthcare ambitions with its highly profitable cloud computing division.
However, the push toward automation in medicine is not without risks. A Reuters investigation last month detailed instances where physicians encountered inaccurate or confusing responses from consumer-grade AI tools, raising broader concerns about reliability and the need for regulation, even when AI is marketed as an assistive rather than diagnostic tool.
The initiative aligns with Amazon's wider corporate focus on cost reduction and increased automation. The company recently acknowledged layoffs within its robotics division, part of an ongoing series of job cuts. A company spokesperson commented, "We regularly review our organizations to make sure teams are best set up to innovate and deliver for our customers."
Investors reacted positively to the announcement, with shares of Amazon.com Inc. rising roughly 1% during early trading on Thursday. The development underscores the growing investment by major tech firms into specialized, industry-specific AI solutions that move beyond offering mere computing power.
