The App I Had No Business Building

I did not wake up one morning thinking, "You know what I should do? Build a stock trading platform."

I am 50 years old. I do not have a coding background. I do not speak the language software people speak. I still have moments where I look at what is on the screen and think, "I have no idea what any of this means."

But I did have a very clear idea.

I wanted to build something that could help look for short-term opportunities in the stock market, especially around AI and AI-related companies. Not a magic money machine. Not one of those internet promises where someone shows a rented sports car and says you can get rich by Friday. I wanted the opposite of that.

From the very first message, I told Lovable: this system must not promise guaranteed profits. It must include strong risk controls.

That part mattered to me. I have spent enough time around investing to know that people get hurt when tools make things look too easy. A green button that says "go make money" is dangerous if the person pressing it does not understand the risk.

So the origin of MyTradePilot was not really, "Can AI pick stocks?" It was more like, "Can a regular person build a serious tool that helps make better decisions without pretending risk does not exist?"

At first the app was called QuantisAI. Lovable gave it that name in the early build plan. It sounded fancy. Maybe too fancy. Like something a hedge fund would put on a glass door. At that stage, I was just happy anything appeared on the screen at all.

The first version had accounts, watchlists, risk settings, and a scanner. I remember seeing the early pages and feeling this weird mix of pride and suspicion. Pride because I had asked for something and it existed. Suspicion because I did not know if it was real or just a pretty shell.

That became the theme of the whole first month.

Is this real?

Is it actually thinking?

Is it using live information?

Is it protecting the user?

Is it just pretending?

Those are uncomfortable questions, but I think they are the right ones. Especially for someone like me, building with AI instead of traditional coding. The danger is not that AI cannot build things. The danger is that it can build things that look finished before they are finished.

That is why I decided to build this in public. Not because I have all the answers. I absolutely do not. But because there are a lot of people my age who have ideas and assume the door is closed because they cannot code.

I am learning that the door is not closed.

But it does not open by magic either.

You still have to ask hard questions. You still have to push back. You still have to say, "That looks nice, but does it actually work?"

That is where this project began: one big idea, one nervous non-coder, and one very firm rule.

No hype without risk controls.