Learning in the Cloud
What Singapore Taught Me About Scaling Education Without Losing the Humans
Singapore always makes me feel like I’ve walked five minutes into the future.
This trip was no different. Except this time, it wasn’t just the shiny buildings and hawker centres towering above that stole my attention. I found myself sitting in a room, listening to governments, LMS giants, and EdTech experts talking about the exact thing we think about every day at Kai’s Education:
How do we get great learning to every kid (not just the ones with fast Wi-Fi and an iPad?)
The panel I sat in on felt like a live dissection of that question– Thanks to Michael Klemm, from the Singapore Education Network for hosting this amazing event– “How do we upscale education without losing the humans?” someone asked. And I kept thinking, yes, this is the conversation. This is the hot topic.
So here’s the long version. Singapore through the eyes of someone who builds little robots in New Zealand and wants them to matter globally.
Landing in the Land of “Of Course It’s Digital”
Singapore loves systems. You can feel it the moment you tap a card, scan a code, or walk into a community hub.
So many incredible speakers had something to share. Jenn-En Jee from GovtTech Singapore, Phway Pann, AWS, Ryan Lufkin, Canvas and Alex Diaz, Seesaw Learning, all very passionate.
When the panel started talking about designing systems that are built to change, it made perfect sense in a Singapore context. They already live that way.
Phway Pann shared how governments had to “stand up online learning in days” during COVID and how cloud technology made that possible. Not just possible, but at a national scale.
Think 80,000 students taking exams at once, or Indonesia running platforms for millions of teachers and students, all because the infrastructure could stretch without breaking.
That is not a flex. That is a philosophy: agility first, then scale, then future-proof.
And of course, my brain went, right, that is precisely what we solve too, but in K–8 with actual physical robots and kids who sometimes eat the cards.
Centralised Countries, Independent Teachers
One of the best parts of the discussion was the point that many Asian education systems are relatively similar. They have strong curricula and clear directives, but teachers still have to teach real, messy, diverse kids.
Alex from Seesaw Learning, made a point I really liked: The platform was built with flexibility as a non-negotiable, so it can work inside a strict system or a very open one.
That resonated with what we do at Kai’s. We like to take a page out of the teaching book across the globe:
- Texas: aligned, standards-based, show-me-outcomes.
- Finland-style classrooms: inquiry, agency, student voice.
- Special schools: “Today we are learning to wash hands. How do we track that?”
Their “voice and choice” example stuck with me. Sometimes the teacher says, “Respond however you like — audio, video, or text.” Other times it is, “Today is writing day.”
The tech should not argue with the student. It should support. That is good EdTech!
It reminded me why we built screen-free coding first. If a teacher only has Chromebooks twice a week, or students are low-vision, or English is not spoken at home, the learning should still happen. Technology should bend, not the child.
Singapore Is NOT as Uniform as It Looks
Another honest moment came from Jenn-En who basically said, “We look super digital from far away, but actually we still have aunties and uncles who need help, where they have no wi-fi.”
They explained how they went out to ask seniors how they would use digital services, and then kept physical service centres open because not everyone is on an app.
That is such an underrated lesson for EdTech: do not close the old door just because you built a new one. Hybrid is not a bug.
It reminded me again why we continue to design for access. This is why we still create screen-free activities, multilingual and Braille materials, physical mats, and teacher-led lessons. Not everything needs to be AI-first. Sometimes it needs to be print, cut, code.
The AI Part (or Why We Should Not Dump Robots on Five-Year-Olds)
My favourite part was the realistic conversation about AI for younger students. Alex mentioned that there are really two paths: student-facing AI and teacher-facing AI. For K–5, they have chosen teacher-facing.
That is the same choice we made.
Yes, AI can personalise, translate, and adapt. But kids also need to talk to real people, argue with friends, work through frustration, and learn to debug.
We can absolutely use AI to make lesson development faster or to translate content into local languages, but the teacher should always be the driver, not the bot.
Ryan from Canvas also made a point that I keep hearing in U.S. schools: AI is forcing universities to rethink assessment. Not because it is fashionable, but because students can now cheat faster than the syllabus can update.
The solution is more authentic work, more skills-based outcomes, and less surveillance. That is a big shift.
What This Means for Kai’s Education
As I listened, I kept connecting the dots back to our work.
Cloud + Edge = Rural Kids Get to Join the Party
Indonesia is already doing low-latency, localised learning in rural areas. That shows our approach — physical robot, optional app, offline activities, sync later — is the right setup for countries with patchy connectivity.
UDL Is Now Mainstream, Not Niche
Multiple speakers talked about universal design for learning as if it were completely standard. That is great to hear. We have been designing for blind students, multilingual learners, and non-readers for years. The world is catching up.
Flexibility Beats Features
No government wants a walled garden anymore. Ryan said Canvas is designed to be open so they can plug into whatever the country needs. That is precisely how we designed our robots. Schools do not want a toy. They want an educational ecosystem.
Policy Moves Slower Than Teachers
Almost everyone on the panel admitted it. Governments set the direction, but teachers adapt to serve kids. Our job is to keep creating tools that teachers can shape, even while we align to TEKS, CSTA, and CASE.
Singapore, the Backdrop
What else did I do. Singapore was its usual mix of efficiency and warmth.
Morning: talking about AI governance.
Lunch: deciding between laksa or chicken rice.
Afternoon: thinking about how to bring KaiBot into a system with national ICT rollouts, Smart Nation goals, and still some digital gaps.
Later that evening, walking past Marina Bay, I thought, this is what future-ready looks like — but only if we remember the child in rural Indonesia sharing one device with three friends.
That is the thought I carried home. You can have perfect infrastructure, but if technology is not designed for mixed-ability, mixed-language, and mixed-connectivity classrooms, it still leaves students out.
The Takeaway I Am Stealing
One line I took back to the team still rings in my head:
“Agility keeps systems responsive, scalability keeps them growing, and future-readiness keeps knowledge alive.”
It is a perfect way to describe what we do at Kai’s Education. We build learning experiences that survive curriculum changes, device changes, and even staff changes.
Our mission is simple: get coding, STEM, and robotics into the hands of kids who might otherwise get left out.
Singapore showed me that the world’s top education systems are thinking about the same challenge. We are just solving it from the kids’ side.
And honestly, that makes me hopeful.
Ronel
Kai’s Education — Future-Ready Classrooms
