The mission - "To help every child understand how AI works" - Simple to say. Genuinely ambitious to deliver.
The moment we sat with it seriously, the hard questions arrived. We are talking about children — so the bar is fundamentally different. This isn't about transferring information. It has to spark curiosity, build intuition, and develop the kind of critical thinking that stays with a child, has to be - Simple enough not to get lost, complex enough to keep the curiosity.
Who delivers it — a platform or a person? How does a child engage with it — by watching or by doing? Where does it belong to — a child's free time or the school curriculum? How do we reach every child — and not just make something available, but make it genuinely accessible?
This is the story of that translation. And how we tried to make sure nothing important got lost.
The first instinct — and why we moved past it
When we shared the mission, the suggestions came quickly. Weekend workshops. Summer camps. An online course on Coursera. A book.
We considered each of them seriously. But each one hit a wall when held against the mission.
Weekend workshops and summer camps could deliver the experience we wanted — teacher-led and interactive. But they don't scale. You can run a workshop for thirty children. You cannot run one for every child. The model is fundamentally limited by time, geography, and the number of trained facilitators you can put in a room.
Coursera and platforms like it solve the scale problem — but create a different one. The kind of custom, interactive activities needed for children to learn simply could not be built inside someone else's platform.
A book ruled itself out on the same grounds. AI moves too fast for print cycles. And no child ever understood something by reading about it that they needed to experience first-hand.
Each option solved part of the problem. None solved all of it.
That gap is what forced us to build our own.
Decision 1 — It had to be our own web app. With zero friction.
If reach was the goal, the delivery method had to match. A web application was the obvious answer for reach. But accessibility meant more than just being online. A child should be able to open a browser, login and begin — nothing to install, nothing to configure, nothing standing between them and the first activity.
And the experience had to live up to the mission. The platform had to support structured courses — each built around lessons that progress intuitively, with hands-on activities woven through every step. Not a one-off experience, but a growing library of learning that schools and teachers could build on over time.
Decision 2 — Hands-on, not just watched.
Our teaching philosophy demanded more than watching and listening. Every concept had to be experienced — through interactive activities conveying core principles. The platform had to support that without constraint.
That ruled out platforms like Coursera and Udemy immediately. They are content delivery systems, built around video lectures. While they support activities, they don't offer the level of customization we needed — purpose-built activities where a child could touch, manipulate, and break simply cannot be built inside someone else's platform.
That required a platform built to support interactive hands-on activities, videos, pop-up quizzes etc.
Decision 3 — The teacher is central to AI literacy
Self-paced independent learning was a deliberate design choice — a child sitting with a parent, a school recommending it informally, a curious teenager exploring on their own. Reachability demanded that the platform worked in all of these contexts.
But we held an equally strong belief about the role of teachers. At this age, children look up to their teachers. AI literacy — something new, something that shapes how they think — is exactly the kind of subject that benefits from a teacher's guidance, context, and conversation. A platform that bypassed that relationship would be missing something important.
This mattered for reachability too. In rural and underserved areas, where platform access to all students isn't guaranteed, the teacher often remains the only consistent point of contact with learning. If AI literacy was going to travel beyond well-connected classrooms, it would have to travel through teachers.
But when we spoke to schools, the picture was humbling — teachers were as new to this as their students. They needed more than a nudge in the right direction — they needed structured guidance, a baseline of AI literacy, ready-to-use tools, and classroom scripts they could lean on while they found their footing.
These beliefs and findings translated directly into platform requirements. The platform itself had to do more than delivering AI literacy to students . It also had to equip the teachers doing the delivering. That meant structured guidance, lesson plans they could pick up and run with, and enough scaffolding for teachers to confidently deliver AI literacy.
But we also recognise that platform support alone will only go so far. Our commitment to teachers goes beyond the platform — we are genuinely invested in their AI literacy. Not just so they can navigate a tool, but so they understand what AI really is, what it means for the world their students are growing up in. That means running training workshops that go deeper than tool familiarity — building the kind of grounded understanding that a teacher can impart to students.
Train the teachers first. The teachers train the children. That chain of understanding is what makes the learning real.
Decision 4 — Absolute data privacy. Non-negotiable.
This one was never a debate.
We were building a platform for children. A platform that would definitely involve voice, image, and video as part of the learning experience. But we drew the hard line - No personal data collected. No voice stored. No images uploaded to any server. Every AI interaction happens locally, on the child's own device.
This isn't just a privacy policy. It is a design constraint that shaped further technical decisions. And we believe it's the only responsible way to build AI tools for children.
Decision 5 — Built to live inside the school day.
Teacher support was one layer. The other was the school itself.
For AI literacy to belong to the curriculum — not sit alongside it as an optional extra — schools needed more than a platform they could point students towards. We felt that the Curriculum designers need to be able to map or create courses \ lessons to their existing syllabus. Creative teachers should have the freedom to sequence lessons their own way, customize activities for their classroom context, and bring their own flair to how concepts are introduced and explored.
We wanted to build that kind of capability in the platform - because a platform that only works one way will always be someone else's course. A platform that adapts to the school becomes part of how that school teaches.
That is the platform we set out to build.
In the next post, we'll get into the thoughts behind our flagship course - ‘Getting to know AI’ — Why the course is structured the way it is!