UnitedHealthcare has launched ‘Avery,’ a generative AI-powered companion embedded within its app and myuhc.com, marking a shift from static chatbots to what the company describes as more agentic, action-oriented healthcare support.
The tool is already live for 6.5 million employer-sponsored members and 160,000 Medicare Advantage beneficiaries, with plans to scale rapidly to 20.5 million members across commercial, Medicare, and Medicaid populations by year-end.
Unlike traditional payer chatbots—often limited to answering basic administrative queries—Avery is designed to actively coordinate care. When a member needs to see a doctor, the system can interpret their benefits, compare provider options based on cost, and take the next step of contacting in-network providers to schedule appointments directly. This moves the experience from passive information retrieval to execution.
The impact on operations is already notable. UnitedHealthcare reports that 90% of member issues are resolved without escalation to a human representative when Avery is used, suggesting meaningful reductions in call centre demand and administrative friction.
“People want health care to be easier to use and tailored to their personal needs,” said Dan Kueter, CEO for UnitedHealthcare’s commercial business. “Avery is one way we are responding to consumer demand for a more coordinated and simpler experience.”
For cases requiring human intervention, Avery enables a more efficient handoff. The system automatically compiles a structured summary of the member’s query, history, and any uploaded documentation before transferring to a live advocate. This reduces repetition for the member and allows support staff to focus on resolution rather than information gathering.
The launch reflects broader investment priorities at UnitedHealth Group, which is allocating $1.6 billion to AI initiatives this year. As payers look to reduce administrative burden while improving member experience, tools like Avery signal a move toward AI as a frontline coordinator—rather than just a support layer.
Click here for the original news story.