Specialty medical practices may be the first to see real value from agentic AI, according to John Beck, senior vice president of strategy at Nextech, an electronic health record vendor focused on fields like ophthalmology, orthopedics, and plastic surgery.
Beck, who previously worked at Allscripts and NextGen, said the complexity of specialty workflows makes them a better fit for AI-driven automation than primary care or hospital settings. The advantage lies in orchestration rather than just insights.
AI vs. agents: The workflow gap
Most AI tools today surface insights, such as flagging overdue claims, but stop short of acting on them. An agent, however, takes the next step by routing those claims into a billing queue, tracking payment timelines, and triggering follow-ups when payments lag. Beck refers to this end-to-end automation as “agentification.”
Specialty practices benefit because their operations depend on teams rather than individual providers. A single patient visit may involve technicians, schedulers, and billing staff across multiple departments. Agentic AI doesn’t replace these roles but redistributes work, allowing each team member to focus on their highest-value tasks.
“A 45% shortage in technicians creates problems,” Beck said. “Agentification doesn’t add staff, but it smooths workflows so existing employees aren’t constantly overwhelmed.”
Related: CMS’ New MSSP Rules Get Thumbs Up
The labor shortage in specialties like dermatology and ophthalmology has pushed practices to rethink operations.
The company’s advantage comes from its deep knowledge of specialty-specific data and workflows. With 20% market share in ophthalmology and 30% in plastic surgery, Nextech identifies patterns that broader EHR vendors or standalone AI firms might overlook. This data shapes how agents are built—not just to automate tasks but to anticipate the unique needs of specialty practices.
A scribe built for specialists
Nextech’s Cora Scribe, an AI-powered documentation assistant, shows the difference between generic and specialty-focused tools. Many EHR vendors offer AI scribing, but Beck describes these as enhanced dictation tools that capture voice input and dump it into charts, often with errors requiring later edits.
Cora Scribe was designed specifically for specialties. It inserts real-time voice input directly into the correct chart sections, matching individual providers’ shortcuts and preferences. In ophthalmology, for example, it can auto-populate detailed exam data without manual entry. The result is a chart completed almost instantly, eliminating after-hours cleanup.
The focus on clinical workflows over revenue cycle tools was intentional. While billing automation is easier to implement, Beck believes reducing administrative burdens on clinicians offers greater benefits. This is especially true in specialties with high documentation demands and severe staffing shortages.
Related: CodaMetrix, Five Health Systems Form AI Coding Council
Automation isn’t always beneficial. Beck cautioned that agents can reinforce bad practices. A tool that automatically rebills unpaid claims after 60 days might seem efficient, but if the claim was denied due to a correctable error, the process becomes counterproductive. “An agent could handle this without human oversight,” he said, “but that’s not wise.”
Consolidation and the churn advantage
For Nextech, this shift presents an opportunity. Many legacy EHR vendors, particularly those relying on client-server models, struggle to adapt to cloud-based, AI-integrated platforms. Beck sees this as a chance to gain market share from incumbents unable to keep pace.
“If a generalist EHR serves specialty markets, behavioral health, and FQHCs, it’s spreading itself too thin,” he said. “Companies must focus on what they do best.”
The move toward agentic AI may speed this trend. Beck predicts agents will soon become common, but their real value depends on implementation. AI isn’t about adding new features—it’s about connecting existing ones more efficiently. That’s a tougher sell for vendors treating AI as a standalone product rather than an operational tool.
Specialty practices face a clear challenge: integrating agentic AI into workflows already strained by labor shortages. Those that succeed will use agents to maximize limited resources—not just automate tasks but rethink how work flows through their teams. Quality access improvements often start with these operational changes.
