I Got This

For sprint 4, I decided to build a routine app for my preschool kids.

The morning chaos

Getting my kids ready for school had been brutal. They both cling onto me for attention and throw tantrums at every opportunity. Potty, wash up, take off PJs, put on fresh clothes, help where needed, breakfast, pack lunch, snack, water, brush teeth, wash face, head out the door. Every single morning.

I know they’re not bad kids. Maybe they simply don’t know what to do. I’ve always known routine charts can help. I’d made some versions before, but they always took forever to source the right images and still missed something. Making something durable and versatile (printing, laminating, magnets or Velcro) takes real dedication. I have a half-finished routine chart sitting around that never got done because I wasn’t sure it would get used after all that effort.

On top of that, I only had one more school week before summer break. If I was going to test this, it had to be now.

Testing the concept first

My biggest risk was spending precious time on something the kids wouldn’t use. So I built a quick prototype with Claude and tried it the very next morning.

Wow. I haven’t felt this peace before. My kids were eager to complete each task so they could tap the buttons. We often skip brushing teeth or washing faces in the rush to get to school on time. But that morning, we did them all and still got to school early.

Building the real thing

It worked, so I expanded the project from scratch: multi-child support, redesigned data hierarchy, editable on the fly but child-proof so only adults can change it.

The hardest part of making a routine chart for kids who can’t read is the pictures. While pictures are helpful, real photos are even better. But sourcing, printing, and laminating them is so troublesome that most charts use generic icons or images. So I added the ability to snap a photo or use one from your library, right in the app. The thing that made physical charts impractical is what makes the app worth it.

One thing I want to note about building with AI. Vibecoding is often touted as a tool that lets anyone develop anything, but from my experience across a few projects now, it’s its own skill to master. It’s not simply improving your prompts; you need to understand the code structure too.

And you need experts around you. When AI started coding in pure JavaScript and DOM API, I didn’t think twice about it. Then my husband, who leaned over to check out what I was doing, asked a few questions and suggested I use Vite and React instead. That one comment changed the whole project. It made me study why certain tools are better suited than others. It’s a reminder of how much experience and expertise matter, even in fields people say AI is replacing. It’s time for us to buckle up and study even more than before.

From a non-programmer trying to build things using AI: pick a few tools and go deep, not broad. The goal is to make good things, not to become a developer in the traditional sense.

Check it out here: I Got This

Wrapping up

I’m finishing this sprint in one week instead of the usual two. Summer break starts, so the morning stress goes away, and the app is in good enough shape that I can create other routines on the go if needed.

In sprint 3, I worked through the Mother’s Rule and the five P’s to understand what I’m actually responsible for as a mother. Sprint 4 feels like a natural segue to build something in the overlap: a Build project for me (Person) that directly serves my kids (Parent) by empowering them to take care of themselves. Given the pain point, I might have still built the routine app for preschoolers anyway, but the framework gave clarity to where things fall and gave me peace.