The hype cycle gets this backwards
Every week there’s a new piece about how AI is transforming leadership. How it’s making executives faster, leaner, more decisive. How the COO role is being reinvented by a handful of prompts and a subscription.
I want to offer a different take, one grounded in what I’ve actually experienced building a regulated SaMD company with AI embedded in our operations.
AI didn’t make me a better COO. I already knew how to operate.
What it did was something more interesting: it made the gap between disciplined operators and everyone else impossible to ignore.
What AI actually rewards
There’s a pattern I’ve observed, both inside Newel and in conversations with other operators across digital health and medtech. When teams adopt AI without a foundation of operational clarity, things don’t get faster, they get noisier. Decisions get made more quickly, but they’re still the wrong decisions. Processes get automated, but the underlying logic was never sound to begin with.
AI doesn’t manufacture clarity. It multiplies whatever is already there.
If your processes are well-defined, your decision rights are clear, and your team has a shared understanding of what good looks like, AI becomes extraordinary leverage. Every task it touches compounds your existing discipline.
If those foundations aren’t in place, AI accelerates the chaos.
This isn’t a critique of the tools. It’s a statement about what the tools require of you before they can deliver on their promise.
What building in a regulated environment taught us
At Newel, we develop AI-driven Software as a Medical Device, products that go through rigorous MDR regulatory pathways and sit inside quality management systems certified to ISO 13485. Vagueness isn’t just operationally inefficient in our context. It’s a compliance risk.
That constraint, which might sound like a limitation, has been one of our greatest operational assets.
Years of building in a regulated environment forces a kind of thinking discipline that most startups never develop. You learn to document decisions, not just make them. You learn to define acceptance criteria before you build, not after. You learn that “good enough” is a phrase that has no place in a quality management system, and by extension, in any serious operation.
When AI entered our workflow, it landed on that foundation. And the results were compounding in a way I hadn’t fully anticipated.
Where AI creates real leverage for a COO
Let me be specific, because I think the conversation around AI and leadership stays too abstract to be useful.
The areas where I’ve found genuine, compounding leverage as a COO are not the obvious ones. It’s not about generating reports faster or summarising meeting notes. Those are productivity gains, not strategic ones.
The real leverage shows up in three places.
First, in decision preparation. AI has become an exceptional thinking partner for stress-testing decisions before they’re made, surfacing assumptions I haven’t challenged, identifying second-order consequences, forcing articulation of the criteria I’m actually using. This doesn’t replace judgment. It sharpens it.
Second, in cross-functional alignment. One of the hardest parts of a COO role in a scaling startup is keeping regulatory, clinical, engineering, and commercial teams operating from the same picture of reality. AI-assisted synthesis of information across these domains, regulatory landscape, competitive signals, operational metrics, has meaningfully reduced the time between “we have information” and “we’ve made sense of it.”
Third, in quality and compliance operations. This is specific to our industry, but the ability to use AI to support documentation, traceability, and review processes within our QMS has been significant. Not as a replacement for human oversight (that’s non-negotiable) but as a force multiplier on the rigor we were already applying.
The question every COO should ask before adopting AI
Before asking what AI can do for your operations, ask whether your operations are worth accelerating.
It’s a harder question than it sounds. It requires honest assessment of whether your processes are actually designed, or just habitual. Whether your team has clarity on decisions, or just the illusion of it. Whether you’re building on a foundation, or on assumptions you’ve never tested.
If the answer is yes, your operations are solid, your team is aligned, your decision-making is disciplined, then AI is one of the most powerful tools available to a modern COO. Move fast and use it aggressively.
If the answer is not yet, fix that first. No tool changes that math.
What this means for digital health specifically
The SaMD space is entering a period where AI will become a standard component of both the products we build and the operations behind them. Regulatory bodies are actively developing frameworks for AI-enabled medical devices. Pharma and medtech partners are evaluating vendors not just on what they’ve built, but on whether their infrastructure can sustain AI-driven development responsibly.
The companies that will lead in this environment aren’t the ones moving fastest in isolation. They’re the ones who built compliance, quality, and operational discipline into their DNA early, and are now able to move fast because of that foundation, not despite it.
At Newel, that’s the bet we made from day one. H.Core, our adaptive development platform, was built to integrate regulatory compliance, behavioral science, and AI in a single operating environment, not as separate workstreams to reconcile later.
That’s what lets us move at startup speed while meeting the standards a medical device company must meet.
AI didn’t create that. It rewards it.
