New Jersey-based Redefine Healthcare, an orthopedic and spine practice with more than 50 office locations and four ambulatory surgery centers, provides a good example of how agentic AI is starting to reshape revenue cycle management at large physician practices.

Redefine Healthcare execs are using AI orchestration capabilities from a company called Adonis to monitor their revenue cycle data, alert the team to high-impact issues, and autonomously deploy agents to resolve issues. 

Using AI agents, Redefine Healthcare said it is reducing denials, underpayments, and downcoding, while improving workflow efficiency through the autonomous resolution of revenue cycle tasks, such as prior authorization workflows, claim status inquiries, and appeals processing.

Kathy Saunders, senior vice president of revenue cycle management at Redefine Healthcare, spoke with Healthcare Innovation about working with Adonis. 

“Adonis highlights issues that they're seeing, and they’re able to summarize it for us on an alert screen, showing us if there are trends,” she said. “If one of the payers is paying us lower, if there's a very specific CPT code that's getting denied for whatever reason, it's highlighted to us, so my team has a better chance of fixing it immediately, and also stopping these problems.”

Redefine uses athena’s EHR and revenue cycle management solution. Saunders said the Adonis solution is able to “highlight Horizon NJ Health underpaying us on this particular code, and then right in Adonis you can work your claim. It links directly to the claim in athena, so if we have to do a resubmission or an appeal while you're in Adonis, it's bringing you right into the athena claim, so there's no stumbling back and forth between the two software programs.”

Saunders added that the real key has been the ability to find these trends and fix them so they don't happen again. “Our turnaround time for reimbursement is much quicker. Adonis has uploaded all of our contract rates pursuant to every payer,” she explained. “They're watching every payment that's coming in, and now we have an underpayment link in Adonis showing us when any of our claims are being underpaid. That has been a game-changer for us.”

Akash Magoon, Adonis CEO, said he and his co-founders have been building Adonis for the last four years and have raised $95 million in total. He said the company’s customer base includes large physician practices such as Redefine Healthcare as well as some of the nation's largest hospital systems, including Mount Sinai Health System.

I asked him about whether there is an AI arms race going on between payers and providers. “I've personally changed my stance on this ever so slightly over the last year,” Magoon responded. “I think that the initial stance was that the providers and the payers are complete enemies and the providers want to get paid and the payers don't want to pay out. I think there's still a little bit of that, but at the same time when a provider organization like Kathy’s has to submit a bunch of appeals and denial rejections, that adds administrative burden on the payer side as well. So I do think a lot of the forward-looking payers are interested in removing a lot of the waste in the system because they realize that it costs them a lot of time, energy, and resources.”

One example, he said, is that Adonis built a voice AI layer that calls insurance companies tens of thousands of times on a weekly basis on behalf of customers. “That’s something that Kathy's team used to have to do manually, and now we do in a very automated-first fashion, but if you think about it, the insurance company also has a massive call center, and they're fielding these calls,” Magoon said. “In a perfect world, none of these calls would actually have to happen. So I think with the advent of AI making things more automated on both sides, there are a lot of costs that can be removed, and I think over time the payers and providers ultimately want to find the best way of being able to work with one another to actually facilitate the interaction.”

Saunders also weighed in on the idea of an AI arms race. “Do I feel like we go head to head — our AI against their AI? Yes, I do at times," she said. “But I do find that AI has helped us considerably to be a better practice, better billers, better coders, making sure that our claims are going out cleaner and catching items that we normally wouldn't catch, or it would take us months to catch. We're catching it much quicker now, so that's allowing us to have that first claim go out clean, get paid much quicker.”

The Adonis AI agents have been a huge help for her team, she said, because phone calls are burdensome. “They’re getting through way more phone calls than any of my humans ever could, and the results are tremendous,” Saunders said. “They come back with very detailed reporting reference numbers — who they spoke to, what the claim needs — and it's coming back to my team for them to apply the action. Once AI gets to the point where they can actually do the action, that’s going to be the next game-changer for us.”