2026
MOMO
An educational app where kids learn by playing, choose their own path, and even build their own questions with an AI companion.
AI Design
Product Design
Design Systems
Product Overview
“Kids don’t need to be told they’re studying. Give them a game, a companion, and a real choice, and they learn without ever noticing.”
Learning that feels like play, not homework
MOMO is an educational app for children that turns the topics they are meant to learn at their age into a game. Instead of being walked through a lesson, kids play, pick what interests them, and build their own path, absorbing the knowledge without realising they are studying.
I built MOMO on my own, in parallel with the AI Designer Pro program at HeroCamp, as a live testing ground for everything I was learning about designing AI products and designing with AI. It was never an exercise. It is a working prototype, now in beta, made with real kids.
Learning by Playing
The point was never to hide the learning. It was to make the playing the reason they learn.
MOMO starts from how children actually engage. They commit to what they choose and enjoy, not to what is imposed on them. So the design hands them agency: they decide which topics to explore and how far to push, and the learning rides along inside the game loop. Motivation comes first, and the content follows it.
Built With Kids
MOMO wasn’t designed for kids. It was designed with them.
The game, its questions and its monsters were co-created with two children, aged 8 and 10, and then tested with five more. From the first version they were co-authors, not just testers. I built, put it in their hands, watched what actually happened, and iterated on what I saw, again and again. Designing for a young audience means you cannot assume anything, you have to observe, and that is what shaped every decision, from the wording of a question to how the game feels when you get one wrong.
How It Works
A simple, honest core: play, answer, progress, unlock.
The current version is a quiz game built around questions with four possible answers. It adapts to the child: you set the age, from first readers around 6 or 7 up to about 12, and the game meets them there. Answering well moves them forward, and progression unlocks new monsters. The monsters are not just rewards. Each one is the face of the AI that guides the adventure, and kids pick whichever companion they want to travel with.
An AI Made for Kids
The interesting part isn’t that MOMO uses AI. It’s that kids use the AI to build the game.
MOMO ships with a set of designed questions, but a child can also create new ones with an AI adapted for kids, and what they make joins the app’s shared pool. So a child is not only choosing what to learn, they are adding to the game for every other player.
The companion does more than ask questions. It shares little facts, reacts to how the child is doing, and prompts playful moments, like asking them to imitate a sound out loud, so learning feels like playing with a friend rather than being tested.
The AI that makes this possible has a face: the monster the child picked. Designing it, how it speaks to a child, what it should and shouldn’t do, and the parental controls around it, was the heart of what I set out to learn.
How I Built It
MOMO is also a demonstration of how I build now: fast, AI-native, and systematic.
Prototype the raw idea, fast
I start in Lovable, no design system, no polish, just the concept made real enough to react to. That rough, throwaway prototype is exactly what a tool like Lovable is good for.
Turn it into a system
Once the idea holds, I build the design system in Figma with variables and tokens, structured from primitive to semantic to mapped, and Tailwind-ready so it drops straight into AI coding tools.
Build the real app in Claude Code
From there the app gets built properly: structuring the project, designing the AI’s personality, and wiring the connections it needs to work.
Give the monsters life
The monsters were designed by me together with the kids in Figma, then animated in Claude through a system of patterns, puppets and expressions, so each one performs rather than just sits on screen.
Growing It Iteratively
The four-answer quiz is version one, not the destination.
MOMO is in beta and built to grow. The quiz is the first playable core, and the plan is to keep expanding the ways kids can play and learn, one tested iteration at a time, holding to the same rule every cycle: ship something real, put it in front of kids, learn, and go again.
What This Project Was Really About
A kids’ quiz on the surface. Underneath, a study in designing AI for the people who trust it most.
MOMO was how I learned applied AI, by shipping it rather than reading about it. Designing an AI companion for children forces the hard questions early: safety, tone, agency, and how much to let the user shape the experience. It is a beta, and kids already play it. What I really built was a way to turn AI-design ideas into a product small enough to fit in a child’s hands.
More Works
©2026
2026
MOMO
An educational app where kids learn by playing, choose their own path, and even build their own questions with an AI companion.
AI Design
Product Design
Design Systems
Product Overview
“Kids don’t need to be told they’re studying. Give them a game, a companion, and a real choice, and they learn without ever noticing.”
Learning that feels like play, not homework
MOMO is an educational app for children that turns the topics they are meant to learn at their age into a game. Instead of being walked through a lesson, kids play, pick what interests them, and build their own path, absorbing the knowledge without realising they are studying.
I built MOMO on my own, in parallel with the AI Designer Pro program at HeroCamp, as a live testing ground for everything I was learning about designing AI products and designing with AI. It was never an exercise. It is a working prototype, now in beta, made with real kids.
Learning by Playing
The point was never to hide the learning. It was to make the playing the reason they learn.
MOMO starts from how children actually engage. They commit to what they choose and enjoy, not to what is imposed on them. So the design hands them agency: they decide which topics to explore and how far to push, and the learning rides along inside the game loop. Motivation comes first, and the content follows it.
Built With Kids
MOMO wasn’t designed for kids. It was designed with them.
The game, its questions and its monsters were co-created with two children, aged 8 and 10, and then tested with five more. From the first version they were co-authors, not just testers. I built, put it in their hands, watched what actually happened, and iterated on what I saw, again and again. Designing for a young audience means you cannot assume anything, you have to observe, and that is what shaped every decision, from the wording of a question to how the game feels when you get one wrong.
How It Works
A simple, honest core: play, answer, progress, unlock.
The current version is a quiz game built around questions with four possible answers. It adapts to the child: you set the age, from first readers around 6 or 7 up to about 12, and the game meets them there. Answering well moves them forward, and progression unlocks new monsters. The monsters are not just rewards. Each one is the face of the AI that guides the adventure, and kids pick whichever companion they want to travel with.
An AI Made for Kids
The interesting part isn’t that MOMO uses AI. It’s that kids use the AI to build the game.
MOMO ships with a set of designed questions, but a child can also create new ones with an AI adapted for kids, and what they make joins the app’s shared pool. So a child is not only choosing what to learn, they are adding to the game for every other player.
The companion does more than ask questions. It shares little facts, reacts to how the child is doing, and prompts playful moments, like asking them to imitate a sound out loud, so learning feels like playing with a friend rather than being tested.
The AI that makes this possible has a face: the monster the child picked. Designing it, how it speaks to a child, what it should and shouldn’t do, and the parental controls around it, was the heart of what I set out to learn.
How I Built It
MOMO is also a demonstration of how I build now: fast, AI-native, and systematic.
Prototype the raw idea, fast
I start in Lovable, no design system, no polish, just the concept made real enough to react to. That rough, throwaway prototype is exactly what a tool like Lovable is good for.
Turn it into a system
Once the idea holds, I build the design system in Figma with variables and tokens, structured from primitive to semantic to mapped, and Tailwind-ready so it drops straight into AI coding tools.
Build the real app in Claude Code
From there the app gets built properly: structuring the project, designing the AI’s personality, and wiring the connections it needs to work.
Give the monsters life
The monsters were designed by me together with the kids in Figma, then animated in Claude through a system of patterns, puppets and expressions, so each one performs rather than just sits on screen.
Growing It Iteratively
The four-answer quiz is version one, not the destination.
MOMO is in beta and built to grow. The quiz is the first playable core, and the plan is to keep expanding the ways kids can play and learn, one tested iteration at a time, holding to the same rule every cycle: ship something real, put it in front of kids, learn, and go again.
What This Project Was Really About
A kids’ quiz on the surface. Underneath, a study in designing AI for the people who trust it most.
MOMO was how I learned applied AI, by shipping it rather than reading about it. Designing an AI companion for children forces the hard questions early: safety, tone, agency, and how much to let the user shape the experience. It is a beta, and kids already play it. What I really built was a way to turn AI-design ideas into a product small enough to fit in a child’s hands.
More Works
©2026
2026
MOMO
An educational app where kids learn by playing, choose their own path, and even build their own questions with an AI companion.
AI Design
Product Design
Design Systems
Product Overview
“Kids don’t need to be told they’re studying. Give them a game, a companion, and a real choice, and they learn without ever noticing.”
Learning that feels like play, not homework
MOMO is an educational app for children that turns the topics they are meant to learn at their age into a game. Instead of being walked through a lesson, kids play, pick what interests them, and build their own path, absorbing the knowledge without realising they are studying.
I built MOMO on my own, in parallel with the AI Designer Pro program at HeroCamp, as a live testing ground for everything I was learning about designing AI products and designing with AI. It was never an exercise. It is a working prototype, now in beta, made with real kids.
Learning by Playing
The point was never to hide the learning. It was to make the playing the reason they learn.
MOMO starts from how children actually engage. They commit to what they choose and enjoy, not to what is imposed on them. So the design hands them agency: they decide which topics to explore and how far to push, and the learning rides along inside the game loop. Motivation comes first, and the content follows it.
Built With Kids
MOMO wasn’t designed for kids. It was designed with them.
The game, its questions and its monsters were co-created with two children, aged 8 and 10, and then tested with five more. From the first version they were co-authors, not just testers. I built, put it in their hands, watched what actually happened, and iterated on what I saw, again and again. Designing for a young audience means you cannot assume anything, you have to observe, and that is what shaped every decision, from the wording of a question to how the game feels when you get one wrong.
How It Works
A simple, honest core: play, answer, progress, unlock.
The current version is a quiz game built around questions with four possible answers. It adapts to the child: you set the age, from first readers around 6 or 7 up to about 12, and the game meets them there. Answering well moves them forward, and progression unlocks new monsters. The monsters are not just rewards. Each one is the face of the AI that guides the adventure, and kids pick whichever companion they want to travel with.
An AI Made for Kids
The interesting part isn’t that MOMO uses AI. It’s that kids use the AI to build the game.
MOMO ships with a set of designed questions, but a child can also create new ones with an AI adapted for kids, and what they make joins the app’s shared pool. So a child is not only choosing what to learn, they are adding to the game for every other player.
The companion does more than ask questions. It shares little facts, reacts to how the child is doing, and prompts playful moments, like asking them to imitate a sound out loud, so learning feels like playing with a friend rather than being tested.
The AI that makes this possible has a face: the monster the child picked. Designing it, how it speaks to a child, what it should and shouldn’t do, and the parental controls around it, was the heart of what I set out to learn.
How I Built It
MOMO is also a demonstration of how I build now: fast, AI-native, and systematic.
Prototype the raw idea, fast
I start in Lovable, no design system, no polish, just the concept made real enough to react to. That rough, throwaway prototype is exactly what a tool like Lovable is good for.
Turn it into a system
Once the idea holds, I build the design system in Figma with variables and tokens, structured from primitive to semantic to mapped, and Tailwind-ready so it drops straight into AI coding tools.
Build the real app in Claude Code
From there the app gets built properly: structuring the project, designing the AI’s personality, and wiring the connections it needs to work.
Give the monsters life
The monsters were designed by me together with the kids in Figma, then animated in Claude through a system of patterns, puppets and expressions, so each one performs rather than just sits on screen.
Growing It Iteratively
The four-answer quiz is version one, not the destination.
MOMO is in beta and built to grow. The quiz is the first playable core, and the plan is to keep expanding the ways kids can play and learn, one tested iteration at a time, holding to the same rule every cycle: ship something real, put it in front of kids, learn, and go again.
What This Project Was Really About
A kids’ quiz on the surface. Underneath, a study in designing AI for the people who trust it most.
MOMO was how I learned applied AI, by shipping it rather than reading about it. Designing an AI companion for children forces the hard questions early: safety, tone, agency, and how much to let the user shape the experience. It is a beta, and kids already play it. What I really built was a way to turn AI-design ideas into a product small enough to fit in a child’s hands.
More Works
©2026