Jonathan Montoya

I went from -$1,200 in my bank account to 7 figures building AI software. Zero code. Zero experience. The exact roadmap: 1. Pick a boring problem 2. Build it with AI in days 3. Get 1 paying user 4. Scale with content Start small. Build smart. Grow big.

The printer knows when you're already late. I can't prove it. I also won't be taking questions.

If customers keep asking the same question about your offer, improve the explanation. Don't assume everyone needs to read more carefully.

A helpful error message gives a next step. If a practice request for a language teacher is missing information, name the field that needs attention and keep everything else the user entered. Making people restart creates needless frustration.

A ‘quick call’ at 4:58 p.m. has never once understood the meaning of either word.

How to test an idea for a dog walker: Sketch a form that collects date, time window, and walk length. Show it to someone who handles those requests. Ask what information is missing and what they would do next. A conversation can improve the plan before you build.

Six figures used to sound like “we made it.” Now it can sound like “we're okay as long as nothing expensive breaks.”

An example of a useful product test: “Imagine you are helping a bicycle repair shop. Submit a sample repair request, close the page, then come back and find the problem description.” Watch the attempt without giving directions. Confusion is useful feedback.

Ask AI to separate facts, opinions, and guesses in your draft. Then check the facts yourself before publishing.

A good first project teaches you the full loop. Help a language teacher organize practice sessions. Build a small version, test with sample information, show it to a real user, and improve one confusing step. Each pass gives you something specific to learn.

Parents can turn a fifteen-minute errand into a three-location expedition with snacks, negotiations, and one missing shoe.

I love a good deal until it requires buying four things I didn't need.

Build a version small enough that a real person can try it this week. A feature list can hide a lot of unanswered questions.

Turn a vague build request into a clear one. Instead of “make me a business app,” describe a home baker, the task to collect custom cake inquiries, and the fields: event date, serving count, and design notes. Then ask for a proposed flow before code.

The biggest financial head start a child has is something money can’t buy later: TIME. Here’s the hypothetical example in the reel: $1,000 at birth, $500 on each birthday through age 18, and $250 at the end of every month through age 43. Assuming a 10% effective annual return

More room in the budget and more room in the calendar are both worth wanting.

Buying a storage bin feels productive right up until you need storage for the storage bins.

A useful way to collect feedback on a maintenance request tool: Ask someone who handles these requests: “What would stop you using this on a busy day?” That question can reveal missing steps that a general “Do you like it?” will miss.

At what salary would you say someone makes “good money” where you live?

Home maintenance is a subscription with surprise billing dates. The water heater doesn't check whether the car already had a turn this month. What's one ordinary household expense you wish someone had explained before you moved out?

I really want to talk about this one. This is actually pretty cool. So, Magai, so this tool, think of it as having like all the AI tools all in one spot. Okay? Like if you don't want to spend that much money on AI tools, this is actually a really cool tool cuz what it does is,