The next phase of healthtech: agentic AI, virtual-care wins and real interoperability

By MarketScale · June 23, 2026.

A set of announcements produced in mid-2026 signals that the health technology sector has moved past the era of pilot projects and is entering enterprise-scale deployments. Four forces converge at this moment: measurable clinical outcomes in virtual care, new agentic AI integrations in contact centers, Epic Systems' operational expansion, and a federal CMS interoperability framework that is quietly redefining what it means to participate in connected care.

**Virtual observation proves its value with concrete numbers**

The week's strongest case is led by Prime Healthcare, a nonprofit system that operates 55 hospitals in the U.S. The organization has announced the system-wide expansion of Collette Health's virtual observation technology, following a pilot at Paradise Valley Hospital in National City (California) that reduced patient falls in med-surg units by 84%, according to Fierce Healthcare. The scale of the result is large enough to generate institutional confidence: Prime is not only expanding virtual observation, but also announcing the leap toward comprehensive virtual nursing programs.

Holly Miller, CEO of Collette Health, described the partnership as an example of how virtual observation acts as a "strategic foundation for the comprehensive transformation of virtual care." Collette Health reports having worked with more than 185 hospitals across the country and having generated $3.96 billion in savings, according to the May 19 press release cited by Fierce Healthcare. The deal with Prime illustrates a recurring pattern: a strong result in a pilot opens the door to enterprise-wide adoption, without the need for a second validation cycle.

The relevance of this case goes beyond the numbers. Patient falls are one of the most costly adverse events—both in terms of legal liability and direct cost of care—in U.S. hospitals. An 84% reduction in med-surg units is hard to ignore for any clinical operations director. The fact that Prime is moving its entire 55-facility network toward this model, and that it is also expanding it to virtual nursing, indicates that health technology buyers are now seeking not just proofs of concept, but proven scalability.

**Agentic AI enters the healthcare contact center**

On the artificial intelligence front, the agentic AI platform Hyro announced a partnership with Five9, a leading provider of cloud contact centers. According to Fierce Healthcare, the deal makes Hyro the first provider accredited by Five9 specifically for the healthcare sector. The goal is to reduce the time needed to integrate AI agents into a health system's existing contact center infrastructure.

Jess Shea, senior partner director at Five9, noted in the May 20 press release that the deal means health systems "no longer have to choose between a robust contact center platform and vertical-specific agentic AI." This phrase precisely captures the problem the plug-and-play model seeks to solve: until now, integrating conversational AI solutions into healthcare contact centers required costly and slow integration projects that many operators simply could not undertake.

The timing is especially relevant because it coincides with the results of an independent survey, also reported by Fierce Healthcare, which found that U.S. healthcare organizations continue to struggle to move AI from concept to practice. In that context of implementation frustration, turnkey integrations like Hyro and Five9's have immediate appeal for operators under cost pressure and staff shortages.

For the agentic AI ecosystem in particular, this move has important implications. Hyro positions AI agents not as a replacement for the contact center, but as a layer integrated on top of existing infrastructure. This reduces the friction of organizational change and accelerates return on investment. Five9's vertical accreditation model could also become a market standard: a seal that distinguishes AI providers who understand the language, workflows and regulations of the health sector from those who simply offer rebranded generic solutions.

**Epic builds its own operational layer**

Epic Systems, the dominant electronic health record system in U.S. hospitals, is expanding its reach beyond clinical records. The company has launched EpicOps, an operational software initiative whose first application, called Teamwork, focuses on scheduling physicians and staff, as well as resource planning, according to Fierce Healthcare.

The move places Epic in direct competition with providers specializing in healthcare workforce management, a segment that has historically belonged to such specialized solutions. For hospitals that are Epic clients, the appeal is consolidation: fewer relationships with external vendors and tighter data loops between clinical and operational systems. If staff scheduling is natively integrated with the health record, managers can see available capacity against expected care load in real time.

Fierce Healthcare describes the scheduling module as the first move in a longer strategy to build technology operations capabilities within Epic's already established relationships with hospital systems. In practice, Epic already has access to its clients' clinical workflows, patient data and billing systems; adding the operational layer closes a loop that could make it even more costly—and complicated—to replace Epic with any other provider in the future.

**CMS raises the interoperability standard**

Federal policy is providing the structural framework for all this activity. The Centers for Medicare and Medicaid Services (CMS) Health Tech Ecosystem framework establishes voluntary but detailed criteria across four categories: aligned networks, electronic health records and providers, payers and technology partners. The stated goal is to enable real-time, consent-based health data access at scale.

For electronic health records and providers, the framework requires FHIR-based structured data exchange, as well as unstructured clinical documents—notes, PDFs, images—and mandates that appointment and encounter notifications reach CMS Aligned Networks within 24 hours of occurring. Payers must make claims data accessible through Patient and Provider Access API standards.

An optional but significant complement across all categories is support for electronic prior authorization, aligned with HL7 FHIR Da Vinci standards. CMS frames the framework as voluntary self-attestation, but organizations that do not meet the full criteria of a given category are directed to commit as a "Friend of the Ecosystem," a distinction that carries real signaling weight in the market.

Although the framework is nominally voluntary, its market logic makes it almost mandatory for any provider aspiring to work with federally aligned networks. The distinction between "full member" and "Friend of the Ecosystem" will act as a de facto filter that will accelerate consolidation of the sector: those who cannot demonstrate FHIR readiness, 24-hour notification capability and clean claims data exchange will be left off the table in the most relevant tenders.

**Massive data in the service of value-based care**

Aside from the infrastructure debate, Atlantic Health System, based in Morristown (New Jersey), announced a partnership with Garner Health Technology to direct employees of participating companies toward Atlantic physicians, with the goal of reducing costs and improving care quality, according to Fierce Healthcare. Garner's dataset spans more than 60 billion medical records of 320 million patients, according to the May 19 press release.

Saad Ehtisham, president and CEO of Atlantic Health, described the deal as part of the organization's commitment to lead the country in "bending the cost curve for patients and their families." But the deal goes beyond a simple referral engine: Atlantic will also use Garner's data to inform internal quality improvement work, turning the arrangement into a two-way exchange.

In general, these types of deals reflect an industry trend: health systems operating under value-based care models need data at scale to identify which of their own physicians produce the best risk-adjusted outcomes. A dataset of 60 billion records offers a comparative base that an individual system could never build on its own.

**AstraZeneca brings AI to clinical trials**

Operational changes in health systems have a parallel in pharmaceutical R&D. AstraZeneca has set the stated ambition that 90% of its new studies carry a digital health strategy specific to that trial, according to the company's R&D pages. Its Unify platform supports remote, real-time data collection through devices and apps, reducing the requirements for in-person site visits and broadening the diversity of patients able to participate.

AstraZeneca frames digital health as central to compressing drug development timelines, arguing that continuous collection of patient data at home can generate faster endpoints than traditional periodic site visits. The approach spans its oncology and biopharmaceutical portfolios, which include cardiovascular, renal and metabolic, and respiratory and immunological programs.

In general, the integration of digital health technology into clinical trials is gaining traction throughout the pharmaceutical industry, driven both by the need to accelerate time to market and by regulatory pressure to improve the diversity and representativeness of trial participants.

**What it means for sector operators**

Taken together, the week's news describes a sector in which tolerance for unproven pilots is waning. Buyers of virtual care, AI and data tools are demanding documented results—fall rates, savings figures, integration timelines—before committing to enterprise deployments. Evidence is no longer optional; it is the price of entry.

The CMS interoperability framework adds a compliance dimension that will likely accelerate provider sorting. Organizations that can demonstrate FHIR readiness, 24-hour notification capability and clean claims data exchange will be positioned differently from those that cannot. For healthcare IT providers, meeting that standard is no longer optional if they want a seat at the table with federally aligned networks.

For agentic AI developers and investors, the Hyro and Five9 move marks a path others will follow: vertical-specific accreditation as a mechanism for differentiation and reduced adoption friction. AI agents that can integrate into existing infrastructure—contact centers, scheduling systems, clinical trial platforms—without requiring custom integration projects will have a significant competitive advantage over those that demand replacement of existing infrastructure.

The convergence of all these moves points to a healthtech that has entered a phase of selective maturity: not all technology matures at the same time, but in virtualization, conversational AI and data interoperability, the clock has gone from measuring quarters to measuring weeks.