The First Wall Felt Huge

When I started MyUPP, I knew it was ambitious.

A Universal Provider Passport is not a little weekend app. The idea is to help physicians carry one trusted credentialing profile, so hospitals and payers are not starting from scratch every time. In plain English: take a process that can drag on for 90 to 180 days and try to shrink it to days.

Even writing that sentence makes me want to lie down for a minute.

I am 50. I do not have a coding background. I am building this with an AI app builder, which is both amazing and occasionally humbling in a very personal way.

The first big wall I hit was not the color palette, although I did spend time choosing the “Clinical Trust” look — deep navy, medical teal, the whole serious healthcare feel.

The wall was this:

This was not one app. It was several apps pretending to be one.

There was a Provider portal. A Hospital portal. A Payer portal. A Credentialing Specialist portal. Each one needed to see different things. Each one needed to do different things. And none of them could accidentally see the wrong information.

That last part matters a lot in healthcare.

At first, I was thinking like a regular person: “Well, people log in, and then they go to their dashboard.” Simple enough.

Then I started understanding what that really meant.

A physician should see their own passport. A hospital should see what it is allowed to review. A payer should not wander into hospital information. A credentialing specialist may need broader access, but not unlimited chaos. Every button, every page, every saved piece of data had to respect who the person was and what they were allowed to see.

That is where I met one of those terms that sounds like it belongs in a server room: RLS.

It stands for Row Level Security. I had to look at it in ordinary language: rules that decide which rows of information a person is allowed to see or change. Like saying, “You can open this filing cabinet drawer, but not that one.”

The concept made sense.

Making it real inside an actual backend was another matter.

The app was connected to Supabase, which is where the real data lives. Not pretend data. Not just a picture of an app. A real backend with login, accounts, tables, permissions, and all the things I had spent most of my life happily not thinking about.

I had chosen email and password login, plus Google login. That sounded normal and modern. But once people can log in, the next question is much harder: what exactly are they allowed to do after they get inside?

That was the moment I felt over my head.

Not “this is slightly tricky” over my head. More like “I may have accidentally walked into the cockpit of an airplane” over my head.

I had to narrow the first version. The full dream included everything: all the portals, document OCR later on, a marketing landing page, and the bigger national platform idea. But the first useful version had to become smaller: Provider Passport plus the Credentialing Engine.

Even that did not feel small.

I remember the emotional shift more than any technical detail. At the beginning, AI made everything feel possible. Then the first wall reminded me that possible is not the same as easy.

The builder can generate screens. It can help structure things. It can explain. But it does not remove the responsibility of deciding what the app actually is, who it serves first, and what must be protected.

That was a sobering moment.

It also helped.

Because the wall forced me to stop trying to build the whole national healthcare system in one pass. It made me ask a better question: what is the smallest real version that still honors the purpose?

For MyUPP, that meant starting with the physician’s passport and the credentialing flow around it. Not everything. Not yet.

I did not feel brilliant. I felt humbled, a little frustrated, and very aware that I was learning in public.

But I did not quit.

Small takeaway: the first wall may be where the real project finally introduces itself.