The Day I Made It Smaller

When I first started MyUPP on 2026-06-06, I had one of those ideas that feels clear in your head and enormous the second you try to build it.

The full name is Universal Provider Passport. The goal is simple to say: help cut physician credentialing from 90–180 days down to days.

Simple to say. Not simple to build.

In my head, this was going to be a national platform with different doors for different people. A provider portal. A hospital portal. A payer portal. A credentialing specialist portal. Each group would sign in, see their own information, move documents and approvals along, and somehow it would all feel calm and trustworthy.

I could picture it. That was the dangerous part.

Because picturing software and building software are two very different hobbies.

I am 50 years old. I do not have a coding background. I am building with AI, which is amazing, but it does not magically remove the need to make decisions. In some ways, it gives you more rope. You can ask it to build almost anything, and it will cheerfully start helping you build too much.

That is what happened here.

At first, I wanted all the portals. All the workflows. All the polish. I wanted email and password login, Google login, a real database, security rules so people only saw what they were supposed to see, and a design that felt serious enough for healthcare.

I even had a visual direction: Clinical Trust. Deep navy. Medical teal. Calm, credible, not flashy.

That part felt good. It made the idea feel real.

Then the weight of the whole thing started to show up.

Not in one dramatic crash. More like a slow realization: every portal was its own project. Every user type meant new screens, new permissions, new places where something could go wrong. A “credentialing platform” is not one app. It is a small city.

And I was trying to build the city before I had built one sturdy street.

The trigger for changing direction was seeing the scope laid out in plain terms. Provider, hospital, payer, credentialing specialist. Real backend. Login. Security. Healthcare-level seriousness. Later, document OCR, which means having the app read documents. A marketing landing page. More features. More layers.

I remember thinking, not in fancy business language but in ordinary panic: this is too much.

That was the moment the direction changed.

I did not abandon the big idea. I narrowed the first version.

V1 became the Provider Passport + Credentialing Engine.

In normal language, that means: start with the doctor’s core credentialing profile and the part of the system that helps organize and move that credentialing process forward. Not the entire national platform. Not every portal fully built out. Not every future feature.

Just the heart of it.

That decision felt both disappointing and relieving.

Disappointing because shrinking an idea can feel like admitting defeat. I had to let go of the fantasy that I could build the whole thing right away just because the AI was fast.

Relieving because the project suddenly had edges. I could see what belonged in the first version and what did not.

Document OCR? Later.

Marketing landing page? Later.

Full multi-portal world? The shell could exist, but the first useful version needed to focus.

This is one of the stranger lessons of building with AI after 50: the hard part is not always getting the machine to do something. Sometimes the hard part is telling it, and yourself, what not to do yet.

I still feel the size of MyUPP. It is a very complex healthcare-grade platform for someone like me. There are days when I look at it and wonder who I think I am.

But making it smaller did not make it less serious.

It made it buildable.

Small takeaway: a smaller first version is not a smaller dream. It is a way to keep going.