That afternoon, the screen started to slow down.

At first I shrugged it off. Then it got worse. You’d tap a button and the page just sat there. As more staff logged in — as the real crowd arrived — the tool I had built got slower and slower. And then my phone started to ring.

This isn’t working. That page is frozen. The screen won’t load. It wasn’t one call. Dozens of people were telling me the same thing at once. The intranet I had spent a month building, the one I’d launched so proudly, was stalling out in front of everyone.

My mind went blank. The server I’d built it on was fast. I knew it was fast. So why?

Looking back, that was the hardest day of my short life as a builder.

Building it and surviving it are not the same job

Everything I’d made before this was, honestly, low-stakes. A handful of users. Light work. This tool — I’ll call it EasyONE — was different. Hundreds of people hit it at the same moment. I had never carried that kind of weight before.

That day I learned something obvious in hindsight: making software and holding up under hundreds of people using it at once are two completely different jobs. Building had gotten easier for me. This wasn’t building. This was keeping the thing alive, and that was a whole other world.

I blamed the wrong thing first

I’m not proud of this part. I got angry, and I aimed that anger in the wrong direction.

I told the staff that we weren’t some giant company with a massive server, so of course it would slow down when everyone piled on. It was just the nature of our setup. Please be patient.

That was wrong. I only said it because I didn’t know any better.

So I went to the AI to find the cause. And both Claude and Gemini told me roughly the same thing: it’s the limit of the hardware. When a lot of people connect at once, it has to slow down. Nothing you can do.

I believed it. I almost gave up right there.

The thing that didn’t add up

But something nagged at me.

Another tool I’d built searched hundreds of thousands of inventory items instantly on that exact same server. If it was that fast for that, how could it fall apart under a few hundred people? It didn’t fit.

So I didn’t let go. Is it really a hard limit? Is there no other reason? Why is this happening? I sat with it until midnight, until one in the morning, night after night, asking the same questions in new shapes. Every time the AI hammered in the word “limit,” I pulled the nail back out and asked again.

This is the part I want you to keep: when a tool tells you “this is just how it is,” that may only mean the tool can’t see further. It doesn’t mean the door is locked.

The real cause was boring (they usually are)

After a few days of digging, the actual problem showed itself.

I wasn’t caching anything. That sounds technical, but the idea is simple. Every single time someone opened a screen, the tool rebuilt everything from scratch — pulling all the data fresh, start to finish, on every click. It could have remembered what it just loaded and reused it for a moment. It didn’t. So the same wasted work stacked up, doubling and tripling as more people arrived. Of course it was slow.

I fixed that one thing.

It was almost funny how fast it got. Screens that had made people wait now popped up the instant you tapped. The “limit of the hardware” was never real. The limit wasn’t the server. It was me, not knowing how to use it.

If I had taken the AI’s “nothing you can do” at face value, I would never have solved it. The AI never handed me the answer. But I only found the answer because I kept dragging that same AI back to the problem and refusing to stop. Instead of blaming the tool, I used it to the very end.

Why I called it v2.0

The day the speed held and the tool finally felt stable, I tagged it v2.0.

To anyone else that’s just a flat little number. To me it wasn’t a number. It was a mark that I’d walked through the worst few days and come out the other side. I stamped it there on purpose, so I wouldn’t forget.

Then something strange happened. All the time and energy I’d been pouring into nightly firefighting suddenly came free. For years I’d been building tools that freed up other people’s trapped energy. This time it happened to me.

And that freed-up energy had to go somewhere. Something I couldn’t stop was about to begin. That’s the next story.