Attention in healthcare is hard-won. And not just in the exam room. Our VP of Marketing Lauren Mitri Abalos makes the case that the attention crisis in healthcare starts upstream, in operations — and the organizations that rethink provider readiness now will have a real operational advantage. Originally published on her LinkedIn, we're sharing it here because the conversation is too important to keep in one place.
Stacy Simpson, CMO of athenahealth, recently wrote something in Fast Company that I loved. Her argument: the future of healthcare isn't about adding more AI capabilities. It's about giving attention back. To clinicians. To patients. To the relationship at the point of care that makes medicine worth practicing in the first place.
She's right. And the conversation she's starting is one the whole industry needs to continue having.
But there's a version of that argument nobody is making yet — and it's the one I think about every day because it's the problem I'm closest to. Before a clinician can give a patient their full attention, someone in a back office has to have done their job. That person, right now, is drowning.
Provider operations teams sit between a health system's growth ambitions and its clinical reality. They manage credentialing, enrollment, onboarding, and provider readiness.
The unglamorous infrastructure that determines whether a newly hired physician can see patients next week or next quarter. Or whether an existing physician stays in-network when a new insurance plan comes on board.
These teams are under extraordinary pressure. Health systems are growing. Provider networks are expanding across states, payers, and facilities. And the process is still, in most organizations, predominantly manual.
Consider what that actually costs. The average credentialing process takes 90 to 120 days to complete.
During that window, a single physician or surgeon loses up to $122,000 in delayed revenue.
At the organizational level, credentialing delays can create $135,000 to over $900,000 in deferred billings over a single 90-to-120-day period, depending on specialty and volume. Multiply that across a health system onboarding hundreds of providers per year, and the math compounds fast.
Beyond rounding errors, these are structural revenue problems hiding inside an operational process that most organizations still run on spreadsheets and email chains.
The attention crisis in healthcare operations isn't talked about in Fast Company articles. But it's just as real and just as solvable.
The broader numbers confirm the scale of the problem. Administrative costs consume roughly 34% of total U.S. healthcare spending. Healthcare professionals report spending more than four hours per day on administrative tasks — time that isn't going to patients, and isn't going to the operational work that gets providers ready to see them.
The industry has started to respond. Administrative AI attracted 60% of all healthcare AI investment in 2024. And, more than half of health plans and a quarter of provider organizations now use AI tools in administrative workflows, according to the 2025 CAQH Index.
But the investment is flowing overwhelmingly toward the clinical side — ambient scribes, documentation tools, EHR automation. The operational infrastructure underneath clinical care is getting a fraction of that attention. And it shows.
Here's where it gets complicated.
The same surge of AI excitement reshaping clinical care is washing over provider operations too. Internal teams are experimenting with LLMs. Vendors are announcing AI-powered everything. And health systems with engineering resources are asking a reasonable question: could we just build this ourselves?
Maybe. But building a workflow tool is a different problem than operationalizing a healthcare workflow responsibly at scale.
Credentialing, enrollment, and provider onboarding are highly regulated, audit-bound processes. They require governance, oversight, expertise, traceability, and compliance infrastructure that doesn't come standard with a general-purpose AI model — no matter how capable that model is.
Consider that among organizations that experience enrollment denials, for example, 40% say application-related errors are a common cause, creating a cycle where unclear requirements and manual steps seem to lead directly to preventable rework.
The value of AI in this environment isn't just speed. It's catching what manual processes miss, consistently, with a defensible audit trail.
You can build the workflow. The harder question is whether you can build the audit trail your NCQA auditor or Joint Commission reviewer will ask for. Whether you can maintain the payer relationship data as enrollment rules change. Whether you can staff the specialist oversight layer that catches the edge cases AI will inevitably surface.
The organizations getting this right aren't the ones with the most AI. They're the ones that paired automation with accountability from the start.
This is what we've learned working with provider operations teams across the country:
It has to be purpose-built. A general-purpose model repurposed for credentialing isn't the same as AI designed specifically for that workflow. The difference shows up in accuracy, in compliance fit, and in how the system handles the edge cases that are actually common in healthcare.
Human oversight isn't optional — it's the feature. The value of AI in regulated workflows isn't that it replaces human judgment. It's that AI handles the repeatable work so credentialing specialists can focus their judgment where it matters: the escalations, the compliance calls, the situations that require a human in the loop. AI should be the second set of eyes, not the replacement for the first.
Accountability has to be contractual, not aspirational. Speed commitments mean nothing without performance guarantees behind them. Health systems deserve (and should demand!) the same contract-backed SLAs from their AI platforms that they expect from any other operational vendor.
The exam room conversation about AI and attention is important. Giving clinicians time back is genuinely worth pursuing.
But health systems that are serious about that goal also need to look upstream. Credentialing and onboarding delays don't just affect revenue. They affect how quickly a health system can respond to patient demand, expand into new markets, and deliver on the clinical capacity they've invested in building. With inpatient utilization projected to rise 9% by 2034 and ED volumes climbing alongside it, the pressure to onboard providers faster is only going to increase.
The organizations that rethink provider readiness now are going to have a fundamentally different operational advantage over the next few years.
Not because they adopted AI. Because they adopted it right.
