Feb 21, 2026 Feb 21, 2026 2026-02-21

When our one-year-old started to watch cartoons on iPad, we mostly used YouTube. He learned very soon how to scroll, how to pick new videos from the “related” column, and, sometimes, he even accidentally subscribed to some channels.

The Accidental UX Researcher

The more he watched, the more a clear UX pattern emerged. He didn’t care about the content — he cared about the interaction. Swipe. Tap. Something happens. Swipe again. New thing. Tap again. Sound comes out. That loop was deeply satisfying for him, and, honestly, it was fascinating to watch. A one-year-old who can’t form full sentences yet, but who had already internalized the core interaction patterns of one of the most used apps on the planet.

And that’s when the question hit me.

So one day I thought: “what if I mimicked the same UX on iPad, but with different content? Instead of the main video, some object image. The related column will scroll to related objects. And instead of Subscribe, we will have buttons to SPEAK that word in Romanian, English, and Vietnamese?”

Three Languages, One Kitchen Table

Now, a bit of context here. Our family is a beautiful linguistic mess. I’m Romanian, my wife is Vietnamese, and we talk to each other in English. So our kid is essentially growing up in a trilingual household, which is both a gift and a daily coordination challenge. At any given moment, the same object on the kitchen table has three names, and all three are correct.

This multilingual reality is what made the idea click. YouTube was teaching him to swipe and tap, but it wasn’t teaching him words — at least not the words we wanted him to learn, in the languages we needed him to hear.

It took me less than half of an afternoon to make the MVP, so we started to test. The first version was rough — just a handful of household items with images, a scrollable list on the right, and three language buttons. That’s it. No animations, no fancy design, just the bones of the idea.

Initially, he was surprised there was no video, but the UX patterns were the same. He scrolled, he clicked, and he tapped the language buttons. Within minutes, he was navigating the app the same way he navigated YouTube. He’d pick an object — say, a chair — see the image, and then tap the Romanian button. “Scaun.” Tap the Vietnamese button. “Ghế.” Tap English. “Chair.” Then scroll to the next object and do it all over again.

That was the moment I knew this wasn’t just a weekend hack. This was something real.

That’s how AI Kiddo was born.

From MVP to App Store

I spent the next few weeks turning the MVP into a proper app. Built it in SwiftUI, kept it lean and focused. No backend, no analytics, no third-party SDKs — this is a kids’ app, and I take that seriously. Everything runs locally on the device. The text-to-speech uses Apple’s built-in AVSpeechSynthesizer, which means it works offline and doesn’t send any data anywhere.

The content is organized into packs. You get three free starter packs right out of the gate — Around the House, Kitchen Essentials, and Bathroom Basics. That’s about 50 objects your kid can explore without paying anything. If they (or you) want more, there are twelve expansion packs covering everything from Animals and Food to Numbers, Colors, Body Parts, Actions, and even Weather. Each pack is $0.99, or you can grab the whole bundle for $2.99.

And here’s the part I’m probably most proud of: it doesn’t just do three languages. It does eight. English, Romanian, Vietnamese, French, Spanish, German, Portuguese, and Korean. You pick which languages you want to display in the settings, and only those show up as buttons. So if your family is Franco-German, you configure it for French and German. If you’re a Korean family living in Spain, you set it to Korean and Spanish. The app adapts to your family’s actual linguistic landscape.

Simple UI Is The Best UI

The UI stays true to the original insight — it still mimics the two-panel layout that my son had already mastered from YouTube. On iPad, you get a split view: the object image takes up the left two-thirds of the screen, and the scrollable list of items sits on the right. On iPhone, it stacks vertically. Big touch targets everywhere — we’re talking 60 to 80 points minimum — because toddler fingers are not exactly precision instruments.

I also added a parental gate, because App Store requires it for kids’ apps, and honestly, it makes sense. Before any purchase or external link, there’s a simple math challenge (something like “15 + 23 = ?”) that a toddler definitely can’t solve. It keeps the experience safe and gives parents control.

One thing I deliberately left out: gamification. No stars, no streaks, no “you did it!” pop-ups. The reward IS the interaction. Tap a button, hear a word. That’s it. Kids don’t need to be tricked into learning — they just need the right tool at the right moment.

Now it’s live on the App Store with 8 integrated languages, 15 content packs, and over 300 vocabulary items. Instead of watching only cartoons, our son is also actively exploring words in multiple languages, building vocabulary at his own pace, driven by the same swipe-and-tap patterns that YouTube accidentally taught him.

Sometimes the best product ideas don’t come from market research or competitor analysis. They come from sitting on the couch, watching a one-year-old accidentally subscribe to a Cars on the Road channel, and thinking: “There has to be a better use for this skill.”

P.S. That’s not my first attempt to solve real-life problems by coding. Stay tuned, there might be another story about how I won against Mekong Delta mosquitoes soon…

If you want to test it, you can download it for free from here: https://apps.apple.com/pt/app/ai-kiddo/id6758517566?l=en-GB