AIdaho
AIdaho — Dispatch No. 05

The Way Through

The frozen practices know the risk. A few figured out how to get the upside without it. Here's what they did differently.

By Will Sundstrom Field interviews July 2026 ~3 min read

Last dispatch was the trap: healthcare leaders who know AI could help, freeze on the risk, and watch their staff use it anyway. This one is about the practices that got unstuck. They didn't ban AI, and they didn't ignore the danger. They did two specific things.

One

Build it inside, not outside

The leaders getting real value stopped reaching for public ChatGPT and built AI inside a secure environment instead. One described her former employer, a national dialysis company, building an internal system that pulled from everything they had:

"It pulls resources from every data point we have in the company. It really is like a business partner."
— Former operator, national dialysis company

Same capability the staff was chasing on free tools. None of the exposure. The difference isn't the model, it's where the data lives.

Two

Aim it at the work, not the people

The practices that got this right framed AI as augmentation, not replacement. It clears the busy work so staff can do the parts that need a human. As one leader put it, the whole thing comes down to "the intent behind the person using it." The wins came from giving people better tools, not from cutting people.

Secure the data, aim it at the busy work. That's the whole playbook.
The Honest Part

Doing it is harder than saying it

A secure environment, a signed BAA, a written policy staff actually follow, someone keeping it maintained as the tools change, that's a lift most practices can't spare a person for. The principle is simple. The build is not.

The Tool That Closes the Gap

Staying ahead of the curve

The company I work with built the solution to exactly this. It's called Metomorph, and it's the reason this dispatch ends with an answer instead of a wish.

Look back at the trap. Staff reaching for public tools because the internal system can't do what they need, and no way to know what patient information went in. Metomorph gives the practice secure AI that's HIPAA compliant, so your team gets the AI they're already using, without the risk. It's not a one-time setup either. It's an ongoing partner that keeps your AI safe and useful as things change. Your data stays yours.

Secure the data, aim it at the busy work. That's the whole playbook.

I'm not going to oversell it. It won't see your patients or make the call a clinician has to make, and it shouldn't. That was the whole point of this series, AI clears the busy work around the care, not the care itself. Which is exactly what every leader I talked to said they wanted.

— W.S., Boise, July 2026

A note on where this comes from

If the freeze sounds familiar.

I'm Will. I just graduated from Boise State and took a job at Ventive, an AI company here in the valley. This series is me trying to understand what's actually happening inside Idaho businesses, and writing down what I find.

If your practice is somewhere in that frozen middle and you want to talk through what getting unstuck actually looks like, my door is open, no pitch required.

Reach out