Adonis AI makes tens of thousands of insurer calls weekly

The revenue cycle — the back-and-forth of billing, denials, appeals and payer rules — has long depended on people investigating denied claims and checking insurer requirements. Adonis is positioning AI agents to take on part of that load, with voice AI already placing tens of thousands of insurer calls each week. But the interviews suggest the technology is the easy part. Salonia Brown of Mount Sinai frames the real work as governance: without operational owners and accountability for exceptions, automation risks "digitizing dysfunction." That theme runs through the other examples. At Rush, frontline staff identify which time-consuming steps add little value, which both guides where to automate and builds staff buy-in. Panoramic Health's Sap Sinha points to a separate constraint: billing, documentation and insurer-interaction rules have not kept pace with the technology, so how those rules evolve could shape whether providers and insurers can use AI agents to interact over claims at all. Adonis, for its part, aims to spot revenue-cycle issues as they emerge and recommend a response — automation in some cases, or flagging an upstream coding or credentialing problem in others. The takeaway for health systems is that adoption hinges less on the model than on the operational plumbing around it.