Boca Raton’s ModMed Takes Aim at Healthcare’s ‘RCM Tax’ With New Agentic AI Platform
Artificial intelligence is rapidly changing how physicians document patient visits and interact with electronic health records. Now, Boca Raton-based ModMed® is turning its attention to another persistent challenge facing medical practices: getting paid.
ModMed announced the ModMed RCM AI Platform, a new agentic artificial intelligence platform built specifically for revenue cycle management (RCM) at specialty medical practices.
The company says the technology represents a shift from using AI primarily as an assistant toward deploying specialized AI agents capable of carrying out portions of complex revenue-cycle workflows.
The goal: reduce what ModMed calls the “RCM Tax™” and the financial and administrative burden created by denied claims, delayed reimbursements, changing payer requirements and the significant amount of human labor required to navigate the healthcare billing process.
For medical practices, the stakes are substantial. ModMed cites an MGMA poll in which 48% of medical group leaders identified denials and appeals as their largest source of revenue leakage. Healthcare systems also spend an estimated $19.7 billion annually appealing denied claims.
From AI Assistant to AI Agent
Much of healthcare’s recent AI adoption has centered on copilots, technology that summarizes information, drafts documentation or provides recommendations to a human user. Agentic AI goes a step further.
Instead of simply providing information, AI agents can be designed to complete multi-step tasks and workflows on behalf of users. ModMed's new platform applies that model across the healthcare revenue cycle, with varying levels of autonomy depending on the task.
The company's AI agents are being developed to perform or assist with activities such as generating appeal packages, suggesting claim edits, validating claims, executing resubmissions, researching payer requirements, navigating payer portals and handling repetitive administrative work. For revenue-cycle teams managing thousands of claims simultaneously, that could fundamentally change how staff spend their time.
Some of those capabilities are already moving beyond recommendations toward autonomous execution. By the end of 2026, ModMed expects to deploy AI agents capable of independently completing more administrative tasks that accelerate RCM workflows, with some capabilities already live.
“For instance, summarizing claims history, suggesting the best next step, or drafting an appeal letter,” ModMed co-CEO Joe Harpaz said. “Human intervention still exists before any sensitive or discretionary areas such as those that would touch coding of claim.”
That distinction helps define where ModMed is drawing the line between automation and human decision-making. AI can prioritize claims, identify likely issues, surface payer requirements, recommend next actions and complete certain administrative steps, while humans remain involved when a workflow requires judgment, carries greater sensitivity or warrants review for accuracy.
Rather than removing revenue-cycle professionals from the process, ModMed is positioning the technology as a way to shift more repetitive and time-intensive work to AI while allowing employees to focus on exceptions, judgment calls and higher-value activities.
“The RCM AI Platform represents a fundamental shift in how specialty practices manage financial performance,” Harpaz said. “By combining specialized AI agents with ModMed’s deep specialty expertise and human oversight, we’ll help practices optimize reimbursements, improve operational efficiency, and give practices greater visibility and control across the revenue cycle.”
That human-in-the-loop architecture could become particularly important as ModMed gives its AI agents greater autonomy. The question is no longer simply whether AI can help an employee determine what to do next, but which portions of the revenue cycle can eventually be entrusted to AI to execute on its own and where human judgment should remain firmly in control.
Five AI Capabilities, One Platform
Rather than introducing a single AI billing tool, ModMed is building the platform around interconnected capabilities.
AI RCM Agents perform or assist with revenue-cycle activities including claim validation, appeals, payer research, claim edits and resubmissions.
AI Workflow Manager serves as the operational command center, prioritizing and routing claims between human employees and AI agents while providing work queues, automation and performance tracking.
AI Billing Assistant gives billing professionals a conversational interface for understanding claim history, submission risks and payer requirements and determining the next best action.
AI RCM Advisor moves further up the organization, translating revenue-cycle data into financial and operational intelligence for practice leaders. It is intended to identify trends, anomalies and opportunities to improve reimbursement and cash flow.
Underlying those technologies is ModMed's Specialty Intelligence Engine, which combines specialty expertise, payer policies, coding guidance, customer workflows, historical claims, denial patterns and real-world revenue-cycle outcomes. That underlying intelligence could prove to be one of ModMed's most significant competitive advantages.
Can AI Prevent Denials Before They Happen?
Perhaps the most significant part of ModMed's strategy is its attempt to move AI earlier in the revenue cycle, from managing denials after they occur to preventing them in the first place.
That strategy is built in part on ModMed's specialty-specific data advantage. The company has spent years developing technology for specialty medical practices across electronic health records, practice management, revenue-cycle management, analytics, patient engagement and payments. Today, its technology is used by nearly 50,000 healthcare providers.
According to ModMed co-CEO Joe Harpaz, that depth of specialty-specific experience gives the company's AI a more granular understanding of how reimbursement challenges differ across specialties and payers.
“ModMed Specialty Intelligence Engine has the ability to understand payer behavior by specialty, very specifically and in near real time,” Harpaz said. “Given ModMed’s breadth of coverage across the country and depth in specialties, there is rarely a denial scenario that ModMed’s AI hasn’t encountered before.”
The advantage, however, isn't limited to the volume of historical claims. Because the RCM AI Platform is integrated into ModMed's broader technology ecosystem, intelligence generated during the revenue cycle can potentially be pushed upstream to the point where an issue originated.
“And updates made in the RCM Platform flow to the practice management system and vice versa seamlessly and in near real time, eliminating the need for data extraction.”
That integration could be particularly important to ModMed's effort to prevent denials rather than simply process them more efficiently after the fact.
According to the company, the RCM AI Platform draws on hundreds of millions of de-identified historical claims, including millions of denied claims, along with billions of dollars in annual transactions, payer-specific intelligence and years of real-world revenue-cycle outcomes. Combined with ModMed's position across clinical, administrative and financial workflows, that data could allow its AI to identify reimbursement risks earlier in the patient journey.
Instead of simply helping a billing employee appeal a denied claim, for example, the platform could identify potential reimbursement issues based on the specialty, procedure, coding, documentation requirements, payer policies and previous claims history before the claim is submitted. That changes the objective from managing denials to preventing them.
ModMed says preliminary internal analysis suggests the platform could reduce manual accounts-receivable follow-up activities by as much as 57%. While that figure is a company estimate rather than an independently validated customer outcome, Harpaz said the projected reduction isn't the result of a single automation. Instead, it reflects potential efficiencies across multiple stages of the denial-management process.
“The reduction in time comes from a combination of prevention recommendations which reduce the number of denials to be worked, in addition to gains throughout the denial management workflow,” Harpaz said.
Among the biggest contributors, he said, are reducing the time employees spend checking claim status, conducting denial root-cause analysis, aggregating paperwork and updating claim fields.
That distinction matters. If AI can prevent some claims from entering the denial workflow altogether while simultaneously reducing the manual work associated with those that do, the potential impact extends across both sides of the revenue cycle equation: fewer problems to resolve and less time required to resolve them.
As payer requirements evolve, ModMed says the platform will adapt its recommendations using updated payer intelligence and practice-specific information. Its integration across the practice-management and RCM environments could also allow lessons learned downstream such as why a particular claim was denied, to inform decisions earlier in the workflow.
The approach reflects a broader shift in enterprise AI: moving beyond chatbots and copilots toward systems capable of coordinating work across multiple applications and executing tasks. In healthcare, however, the consequences of automation make human oversight particularly important.
ModMed's human-in-the-loop model suggests the company sees AI not as a replacement for revenue-cycle teams, but as a way to dramatically increase the amount of work each employee can manage. If the platform can prevent even a portion of denials before they occur, while automating significant portions of the administrative work surrounding those that remain, the value could extend well beyond making the appeals process faster.
South Florida's Growing Healthtech AI Story
Headquartered in the City of Boca Raton, ModMed has grown from a specialty electronic health records company into a healthcare technology platform spanning clinical, operational and financial functions.
Since its founding in Boca Raton in 2010, ModMed has grown from an EHR innovator into a national healthtech company, helping put South Florida on the map for healthcare technology. Its reach has proven durable: by 2024, more than 700 medical practices had been using ModMed technology for at least a decade. Along the way, the company has contributed to South Florida’s innovation economy by attracting talent, investment and national attention to the region.
The company describes the RCM AI Platform as a cornerstone of its broader AI-Powered Practice™ strategy. The strategy extends AI beyond clinical documentation into scheduling, patient communication, eligibility, prior authorization, billing and denial appeals.
“Our data-native foundation has uniquely positioned us to lead the AI charge,” stated Dr. Michael Sherling, co-Founder and Chief Medical and Strategy Officer of ModMed. “We are building technology that is practical, useful and frees up providers and their staff to focus on patient care.”
RCM may prove to be one of the clearest demonstrations of what that strategy can accomplish because its success can ultimately be measured in dollars.
What's Next
Portions of the RCM AI Platform are already being deployed within ModMed RCM Services, where AI is augmenting revenue-cycle operations for the company's managed-services customers.
A limited early-adopter program for customers managing their own revenue-cycle operations is expected to begin in the fourth quarter of 2026, with broader availability planned for 2027. ModMed will publicly showcase the technology at its MOMENTUM customer conference in Orlando from October 16–18.
ModMed plans two adoption models: Larger practices, enterprise healthcare organizations and management services organizations will be able to deploy the platform within their own revenue-cycle teams. Practices using ModMed's managed RCM services will receive the technology as part of a combined AI-and-human service model.
If specialized AI agents can successfully navigate payer requirements, identify reimbursement risks, generate appeals and execute administrative workflows at scale, the impact could extend well beyond faster billing.
For ModMed, the RCM AI Platform is ultimately a test of whether agentic AI can move from helping healthcare workers perform administrative work to actually performing significant portions of that work alongside them and whether doing so can meaningfully reduce one of the hidden costs of running a medical practice.
Boca Raton-based ModMed is taking aim at one of healthcare’s biggest administrative burdens with a new agentic AI platform designed to reduce denials, automate revenue cycle workflows and help specialty medical practices get paid faster.