Agentic AI: The Future of Healthcare Isn't Smarter Software—It's Smarter Workflows

Artificial intelligence has become one of the most discussed technologies in healthcare. Nearly every conference, board meeting, and executive strategy session includes conversations about generative AI, ambient documentation, large language models, and automation. Much of the attention has focused on speeding up individual tasks, such as summarizing patient encounters, generating clinical documentation, retrieving information, or assisting with coding.

These are meaningful advances, but I don't believe they're the most transformative aspect of AI.

The bigger opportunity isn't making clinicians more efficient at interacting with software. It's enabling software to coordinate work on behalf of clinicians.

After spending nearly three decades working in enterprise communications, clinical mobility, and healthcare technology, one lesson has remained remarkably consistent. Healthcare rarely struggles because information doesn't exist. It struggles because information doesn't move.

Islands of technology

Modern hospitals have become incredibly sophisticated organizations. Electronic health records, nurse call platforms, secure messaging, medical devices, workforce management systems, asset tracking, patient monitoring, scheduling applications, and countless departmental systems all perform their individual functions exceptionally well. Yet they often operate as independent islands of technology. Every department has visibility into its own world, but very few systems understand what the organization as a whole is trying to accomplish.

The result is familiar to anyone who has spent time inside a hospital.

Nurses coordinate care across multiple devices while responding to alarms, secure messages, phone calls, and documentation requirements. Physicians move between applications to gather information scattered across different systems. Environmental services waits for notifications. Transport waits for assignments. Pharmacy waits for orders. Case management waits for discharge decisions. Every team performs its role, but much of the coordination still depends on people manually connecting the dots.

From assistants to agents

That is where I believe agentic AI has the potential to fundamentally change healthcare.

Unlike today's AI assistants, which generally respond to prompts, agentic AI is designed around objectives. Instead of answering a question, an AI agent can reason through a goal, determine the steps required to achieve it, interact with multiple systems, communicate with other agents, and complete work within clearly defined governance and human oversight.

The distinction may seem subtle, but it represents a fundamental architectural shift.

Healthcare has spent decades building applications. The next decade may be about building autonomous workflows.

A day in the emergency department

Imagine a patient arriving in the emergency department with symptoms consistent with sepsis. Today, dozens of independent workflows begin almost immediately. Registration captures demographics. Insurance eligibility is verified. Laboratory orders are placed. Nursing performs triage. Physicians begin evaluation. Pharmacy prepares medications. Bed management evaluates capacity. Environmental services prepares rooms. Transport coordinates movement throughout the hospital.

Every one of these activities is supported by technology, yet much of the coordination still relies on clinicians, operators, and administrative staff manually moving information between systems.

Now imagine those same workflows orchestrated by specialized AI agents.

One agent continuously evaluates patient acuity using vital signs, laboratory results, and historical data. Another monitors inpatient capacity and predicts bed availability hours before discharge occurs. Another coordinates communications among physicians, nurses, pharmacy, transport, and environmental services. Another ensures documentation requirements are completed while identifying potential delays before they affect patient flow.

None of these agents makes clinical decisions.

They make operational decisions.

The physician still practices medicine.

The nurse still delivers care.

The AI simply reduces the operational friction surrounding them.

Healthcare's real shortage is time

That distinction is incredibly important because I don't believe healthcare's greatest shortage is intelligence.

It's time.

Every hour clinicians spend searching for information, coordinating logistics, documenting activities, or navigating fragmented systems is an hour they cannot spend with patients. If agentic AI succeeds, it won't be because it replaces caregivers. It will succeed because it gives caregivers greater capacity to do what only humans can.

Beyond interoperability

This also changes how we should think about healthcare technology architecture.

For years, we've focused on interoperability, connecting systems so they can exchange information. While that remains essential, interoperability alone doesn't create coordination. Connected systems still require someone to interpret information and decide what happens next.

Agentic workflows introduce another layer. Instead of simply sharing data, systems begin collaborating toward shared objectives. Communication platforms become execution engines. Clinical mobility platforms become delivery mechanisms for AI-generated workflows. Identity, governance, security, and auditability become just as important as the intelligence driving the agents themselves.

An inflection point

In many ways, healthcare has reached an inflection point similar to previous technology revolutions. Electronic health records digitized patient information. Cloud computing centralized infrastructure. Mobile technologies placed information in clinicians' hands. Agentic AI has the potential to coordinate everything between those systems.

Whether that future arrives in three years or ten remains to be seen. Healthcare moves deliberately, and it should. Patient safety, regulatory compliance, security, and trust will always outweigh the speed of innovation. Human oversight will remain essential, particularly in clinical environments where lives are at stake.

But the direction feels increasingly clear.

The conversation is gradually shifting from "How can AI help me complete this task?" to "How can AI coordinate this entire workflow?"

That's a much bigger question.

The next chapter

Healthcare has always evolved one technology wave at a time. From paper to electronic health records, from pagers to smartphones, from on-premises infrastructure to the cloud, each advancement changed how care was delivered. Agentic AI feels like the next chapter, not because it makes software smarter, but because it has the potential to make healthcare itself more coordinated.

We're still in the early innings, and many questions remain around governance, trust, interoperability, and adoption. But one thing seems increasingly clear: the conversation is shifting from using AI to complete tasks to using AI to orchestrate outcomes.

I believe that's where the real transformation begins.

What are your thoughts? Do you see agentic AI becoming the next major evolution in healthcare operations, or will its impact be more incremental? I'd genuinely enjoy comparing notes with anyone thinking about these questions from the clinical, operational, technology, or investor side.