My First Big Ask

On July 4, 2026, I started what was then called MyAdPilot. Later it became StrategyPilot, and eventually MyBrandPilot.

The name changed because the idea kept getting clearer. Or maybe bigger. Probably both.

My first big prompt for this project was not a small one. I did not ask the AI to make a simple landing page or a cute little tool. I asked for the beginning of an enterprise-style software product with an “elite consulting” feel.

Which is a slightly ridiculous sentence for someone like me to write.

I’m 50. I do not have a coding background. A few years ago, if you had said “enterprise SaaS” to me, I might have nodded politely and then looked it up later. Even now, I translate it in my head as: serious business software that companies might use, not a weekend toy.

But that was the goal.

I wanted MyBrandPilot to feel like something a top consulting firm might use or present. Clean. Structured. Confident. Not flashy for the sake of being flashy. More boardroom than carnival.

I also asked for it to be built in phases. Nine phases, in fact. Again, I can hear the more sensible version of myself whispering, “Maybe start with one?” But at the time, I wanted the AI to understand the whole shape of the thing before we started building pieces.

The first big ask included a lot:

• A polished consulting-style design system
• Organizations and user roles, so different people could have different access
• Website research using Firecrawl, which is a tool that can read and pull information from websites
• A group of AI “specialists” to help generate strategy, instead of one general-purpose answer machine
• A required Critic pass, meaning the AI had to review and challenge the work before treating it as done
• Marketing deliverables that could be generated from the strategy work

That is a mouthful. It was a mouthful then too.

What came back surprised me.

The builder did not laugh me out of the room. It treated the request seriously. It broke the work into a phased plan. It understood the general feeling I was after: not just “make an app,” but make a product with a certain kind of presence.

That was encouraging. More than encouraging, honestly. It was the kind of response that makes you sit up a little straighter and think, “Wait, is this actually possible?”

But there was another feeling underneath it.

The plan was big. Very big. Seeing it organized into phases made it feel real, but it also made it feel heavier. The AI can make ambition look tidy. A nine-phase build sounds elegant when it is written out neatly. Living inside it is another matter.

I remember thinking that I had asked for something that looked simple in a prompt but was not simple at all. Multi-tenant organizations and role-based access sound like tidy business features. In plain English, that means the app needs to know which company you belong to, what you are allowed to see, and what you are allowed to do. That is not decoration. That is plumbing.

And Firecrawl sounded magical, but connecting outside services is always where my stomach tightens a bit. The idea was simple: let the app read a company website and use that information to help create better strategy and marketing outputs. The reality is that “connect this thing to that thing” often becomes where I learn how little I know.

Still, the first response lived up to my expectations in one important way: it gave me a map.

It did not give me a finished product. It did not remove all confusion. It definitely did not make me suddenly feel like a software expert. But it gave the project a structure I could follow, one phase at a time.

That mattered.

Without the AI, I do not think I would have known how to turn a vague idea like “AI brand strategy platform” into a sequence of actual building steps. I would have stayed in the notebook stage, where everything is possible because nothing has to work yet.

With the AI, the dream became more concrete. And once it became concrete, it also became intimidating.

That is the strange bargain of building this way. The AI lowers the front door so people like me can walk in. But once inside, you still find stairs.

Small takeaway: a big prompt can give you a big map, but you still have to walk it one step at a time.