Healthcare Built Digital Memory, Not Intelligence

We built systems of record. Now we need systems that can think, learn, and act.

For more than two decades, healthcare has been on a digitization journey. We transformed paper records into electronic health records and digitized claims, revenue cycle operations, care management, utilization review, patient engagement, and countless administrative processes. We built software, integrated systems, migrated to the cloud, and created an enormous digital footprint across the healthcare enterprise.

Those investments were necessary, as they created something healthcare had never truly had before: digital memory. But digital memory is not the same as intelligence.

Healthcare became incredibly good at storing information. What it never fully solved was how to continuously turn information into coordinated action. Data became more available, decisions remained fragmented, knowledge existed everywhere, yet work still depended on people manually connecting the dots across systems, departments, policies, and workflows.

That’s why I believe we’re entering a fundamentally different era of healthcare transformation. The challenge is no longer information capture. It is now about intelligence orchestration.

The Same Bottleneck, Running Faster

On paper, most healthcare organizations appear digitally mature. They have modern platforms, analytics environments, APIs, cloud infrastructure, and increasingly ambitious AI strategies. Yet, beneath that digital surface, many of the same operational bottlenecks still exist.

Prior authorization is one of the clearest examples and most often being addressed right now due to regulation pressure. The American Medical Association’s 2024 prior authorization survey found that 78% of physicians said prior authorization often or sometimes leads patients to abandon a recommended course of treatment. 

But why? The clinical history exists, the policy exists, the documentation exists, and the expertise exists. 

What doesn’t exist is a reliable mechanism for bringing all of that context together at the moment a decision needs to be made. The result is delays, rework, provider frustration, administrative burden, and patient impact.  And, prior authorization is simply a visible symptom of a much larger issue. The same pattern appears in appeals and grievances, care management, claims operations, payment integrity, quality programs, provider operations, and countless other workflows. 

Information is scattered across systems. Context is reconstructed repeatedly. Teams spend enormous amounts of time finding, validating, interpreting, and transferring knowledge that already exists somewhere else inside the enterprise.

Healthcare built systems of record exceptionally well. What it did not build were systems capable of continuously assembling the right context, logic, and next action around the work being performed.

As a result, organizations often accelerate isolated tasks while leaving the larger system unchanged.

This is why AI deployed into legacy workflows can create the illusion of progress while preserving the same structural inefficiencies underneath. A broken workflow with faster components is still a broken workflow.

The bottleneck simply runs faster.

The Spaces Between Work

Healthcare leaders often focus on optimizing tasks, but healthcare rarely breaks at the task level. It breaks in the spaces between tasks.

It breaks when one team lacks context another team has already discovered or when a clinician has to reconstruct patient history that was already reviewed elsewhere. It breaks when a care manager, claims analyst, pharmacist, and utilization reviewer all touch the same case, but operate from different fragments of the truth.

This is where many automation strategies fall short. Summarization matters, extraction matters, classification matters, but improving isolated activities is not the same as redesigning how work moves through the enterprise. The real opportunity is not task automation, but rather contextual coordination. The ability to continuously assemble the right information, the right expertise, the right policies, the right evidence, and the right next action around the work being performed is what healthcare has been missing.

AI Should Handle the Work Around the Work

The real opportunity is to redesign healthcare work around a more intelligent division of labor. Many healthcare tasks are data-intensive and rules-based, but not necessarily judgment-heavy. Those tasks should not consume the time of clinicians, nurses, care managers, and specialists whose highest value lies in judgment, escalation, empathy, and patient engagement.

This does not mean removing humans from decisions. In healthcare, especially that would be the wrong (and possibly dangerous) approach. Instead, it means designing workflows so humans lead where judgment matters most, while AI supports the administrative and contextual legwork that slows them down. 

AI can gather the evidence, surface the relevant policy, identify what is missing, compare information across sources, flag exceptions, and prepare recommended next steps for review. The human should be able to see the reasoning path, challenge it, override it, and remain accountable for high-stakes decisions.

Healthcare AI cannot be a black box hiding inside an already opaque process. If an organization cannot show where information came from, how it was used, which rule or policy applied, and when a human reviewed it, the workflow is not ready for serious deployment. Auditability is not paperwork after the fact; it has to be part of the architecture.

The Rise of Living Intelligence

The direction healthcare is heading is becoming pretty clear. 

The CMS Interoperability and Prior Authorization Final Rule required impacted payers to meet certain provisions beginning in 2026, with API requirements primarily taking effect in 2027. It is often discussed as a compliance issue, but the broader signal is obvious: healthcare work must become faster, more transparent, and more interoperable.

The same applies on the clinical side. The National Academy of Medicine has identified clerical burden, especially documentation and order entry, as a major driver of clinician burnout. This burden is time lost to typing and attention lost to navigation, with expertise being wasted on system archaeology. We didn’t train highly skilled professionals to be information gatherers. We trained them to make decisions, solve problems, and care for people.

Patients, providers, and payers all expect better experiences. At the same time, healthcare organizations are being asked to do more with less while maintaining trust, compliance, and accountability. That’s why I don’t think healthcare’s next challenge is adding more technology. We already have plenty of technology.

What we need is a different way of operating where the right context is available when and where work happens. Where knowledge doesn’t get trapped inside departments, workflows, or individual teams. Where intelligence builds on itself over time, making every decision a little smarter than the last.

That’s what we mean by Living Intelligence.

It’s not AI as another feature or another application employees have to learn. It’s intelligence woven into the fabric of the enterprise. A continuously learning layer that brings together data, policies, clinical knowledge, operational expertise, and human judgment so organizations can make better decisions, take faster action, and continuously improve.

Most importantly, it should provide faster answers for patients, less time spent hunting for information, more time returned to care planning, and fewer delays, denials, or complaints caused by missing context.

What Healthcare Should be Asking For

If there’s one thing healthcare doesn’t need right now, it’s another AI pilot that generates excitement for a few weeks and then struggles to scale. It doesn’t need another point solution that solves a single problem while creating new silos somewhere else. And it definitely doesn’t need AI bolted onto workflows that were never designed to work intelligently in the first place.

Healthcare already has what it needs: vast amounts of data, sophisticated technology systems, and some of the most knowledgeable professionals in any industry. The challenge is bringing all of those pieces together in a way that helps organizations make better decisions and act on them faster.

The organizations that pull ahead won’t necessarily be the ones with the most data or the largest technology budgets. They’ll be the ones that are best at connecting knowledge, expertise, policies, workflows, and human judgment into coordinated action. They’ll move beyond simply storing information and start building systems that help people work smarter, make decisions faster, and continuously improve over time.

That’s where healthcare transformation is headed, toward a new operating model where intelligence becomes part of how work gets done. Organizations that embrace that shift will be able to adapt faster, scale expertise more effectively, and create better experiences for patients, providers, and employees alike.

In many ways, I believe this transition will be every bit as important as the move from paper to digital records. The last era was about digitizing healthcare. The next era is about making healthcare intelligent.