Healthcare founders, operators, and investors are all wrestling with the same pressure: how do you grow profitably when costs keep rising, margins keep compressing, and the team is already running past it’s limits?
The conversation almost always turns to technology. And these days, of course, that means AI.
The optimism isn’t misplaced. AI genuinely has the capacity to reduce administrative burden, support better clinical decisions, and eliminate the kind of redundant back-office work that has been quietly bleeding healthcare organizations for years. The use cases are real. The ROI potential is real.
But there’s a part of that conversation that always gets overlooked and will determine whether healthcare organizations actually capture that value or just accumulate more expensive tools that don’t deliver.
Before AI can work in healthcare, healthcare must fix its workflows.
The Problem Isn’t Technology. It’s the Foundation Underneath It.
The economics facing most healthcare operators right now are genuinely difficult. Staffing costs are up. Reimbursement pressure is accelerating. Revenue cycle complexity keeps growing. And the organizations trying to solve these challenges with AI are often doing so on top of workflows that were never designed to work together in the first place.
The EHR doesn’t talk to the scheduling system. The billing team has workarounds built on workarounds. Clinicians have 17 steps to complete a task that should take three. Documentation requirements consume hours that should go toward patients.
This is the friction that’s been preventing people from doing their best work — and it’s been normalized for so long that most organizations have stopped seeing it. Research published in JMIR Formative Research put it plainly: AI’s potential in healthcare “remains limited by a complex set of technological, human, and organizational challenges” — including clinician resistance, legacy system constraints, and what the authors describe as the “black box” phenomenon, where clinicians can’t understand or trust what an AI tool is actually doing. The technology itself isn’t the barrier. The environment it’s being dropped into is.
AI Amplifies What’s Already There
Here’s what I’ve come to believe: AI doesn’t remove friction. It amplifies what’s already there. When you layer automation on top of a fragmented, inconsistent process, you don’t get efficiency — you just get to broken faster.
The American Medical Association has been direct about this. Most AI projects in healthcare fail not because the algorithms underperform, but because of structural issues — EHR complexity, physician burnout, and the failure to engage care teams early in the design process. Forbes reporting on the perspective of physician executives echoed the same point: AI tools have to be designed with the clinical professionals who will use them, not deployed on top of them. Human-centered design isn’t a nice-to-have. It’s what determines whether adoption deepens or stalls entirely.
The organizations that are capturing real value from AI right now are not the ones who bought the most sophisticated tools. They’re the ones who did the hard work first. They mapped their workflows. They listened to their frontline teams — not just leadership assumptions about where friction lives, but where the people actually doing the work lose time every single day. They cleaned up their data environments. They standardized before they automated.
That is not a technology project. That is a leadership project.
What Redesigning Work Actually Looks Like
The healthcare organizations that will thrive over the next decade won’t simply adopt more technology. They’ll redesign work. They’ll build smarter workflows. They’ll use AI thoughtfully — not as a shortcut around operational complexity, but as a multiplier on operational clarity.
A recent piece in NEJM AI framed this well: the goal isn’t to replace clinical judgment with automation, but to deploy AI as a “teammate” — something that enhances human capacity rather than circumventing it. That framing matters. It shifts the question from “what can we automate?” to “what does this make possible for the people doing the work?”
This is especially important for founder-led and growth-stage healthcare businesses, where the pressure to move fast is highest and the foundation is often still being built. Scaling a broken workflow is not a growth strategy.
The most important question isn’t “which AI tool should we buy?” It’s “where is the friction that’s preventing our people from doing their best work?” Answer that first. You’ll find that a significant portion of what’s slowing you down doesn’t require AI to fix at all — it requires better process design, clearer ownership, and tighter integration between systems you already have.
Once that foundation is in place, the AI conversation becomes a completely different one. Instead of asking where to plug something in, you’re asking where technology can genuinely accelerate something that already works. That’s when the ROI becomes real, defensible, and durable.
The Future Belongs to Organizations That Become More Productive and More Human
I’ve spent a lot of time thinking about what separates the healthcare organizations that are going to win over the next decade from the ones that are going to struggle — and it’s not access to capital or the latest technology.
It’s a willingness to slow down to speed up. To do the operational work that isn’t on any vendor roadmap. To build workflows that give clinicians, operators, and leaders back the capacity to focus on what humans do best.
Healthcare is under real pressure from rising demand and outdated processes. The potential for AI to help redesign care delivery is genuine. But that potential only materializes when the organizations pursuing it treat workflow transformation as the leadership priority it actually is — not as a prerequisite to check off before buying software.
AI can be a powerful part of that story. But only after the workflow is ready for it.
That’s the work worth doing.
Ryan Kirkpatrick is Managing Director at Rallyday Partners, a private equity firm built by founders, for founders. Rallyday invests in lower-middle market growth companies and brings operational, human, and creative capital alongside financial capital to the organizations it backs.
Sources:
- Forbes: Why Healthcare AI Keeps Falling Short At The Point Of Care
- American Medical Association: Why Most AI Projects Fail in Health Care
- Amplieo: Can AI Fix Healthcare’s Broken Workflows?
- NEJM AI: Redesigning Clinical Workflows with Generative AI
- JMIR Formative Research / NIH: Hype vs Reality in the Integration of Artificial Intelligence in Clinical Workflows
