The Robodonkey Story
We made a gadget for kids that makes asking questions as fun as throwing a frisbee. This is a story of how we found ourselves neck-deep in a quagmire of plastics, circuits, firmware and the peculiar kindness of strangers on the internet.
PART ONE
I used to have a question that haunted me more than the meaning of life: How should a child interact with AI? My kids loved asking ChatGPT all sorts of things — like why cheese is yellow — but handing over my phone felt like giving them a nuclear launch code. I wished there was a way to talk to chatGPT without needing a laptop or phone. Smartphones and laptops come with unrestricted internet access
I thought, "Why not build something just for them?" But there was one hiccup — I had zero clue how to build a gadget. I figured it was time to channel my inner Elon, who usually has the right question ready. Not "What did you do this week?" but rather, "What's the most kid-friendly device ever made?"
As it turns out, there is a device that's been sold 154 million times — possibly the most ergonomic thing ever made for kids.
(Source: wikipedia)
I started seeing a form emerge in my mind. It was exciting. I dove into sketching designs. But then, something terrible happened. My sketches started to resemble ... Netbooks! Yes, those awkward little gadgets that were neither phones nor laptops. They were the mullets of the tech world—business up front, party nowhere.
I hated netbooks, and I hated my sketches. Time for a pivot.
Inspiration struck "Lets go more towards gaming! Not boring, grown-up productivity", I thought. I said farewell to my inner Elon and called upon my inner Jony Ive. But how do you make something "fun"? I had never made anything fun — except that one time when I worked in B2B SaaS. Instead of trying to come up with fun designs, I decided to focus just on posture — specifically, the wrist position. Let me explain: you see, productivity tools force your wrists together, like handcuffs made of tedium. Kids must see this all the time —parents glued to their laptops, typing on their phones, ignoring the world (and them).
I didn’t want my device to be associated with that posture. Gaming, on the other hand, demands wrists apart — wide, free, ready to conquer worlds. Therefore, I figured kids would subconsciously associate this stance with fun. Wrists together meant "work"; wrists apart meant "playtime." I ran with this idea!
I needed a keyboard that kept the wrists apart. But then, a blast from the past hit me — a Nokia E70. Yes, the glorious brick from my youth. It had the perfect design element I was looking for.
What happens when you take the best-selling handheld gaming device of all time and marry it to a Nokia E70? You get a bizarre, yet oddly satisfying vision that just might work. This was my design direction.
My initial sketches were just good enough to convey the idea. So I shared them with an electronics engineer I met on Fiverr. He wasn't just super smart but also a nice person (from Sweden!). He didn’t just help with the electronics, but also taught me a lot along the way — he helped me understand what a PCB is and why ESP32 wasn’t some weird new energy drink.
We made an empty shell first. And when I say "made," I mean we sent it to a 3D printer and crossed our fingers. The excitement of seeing it come out was like watching a baby being born — if babies were made of plastic and took a few hours to print. "It's a start," my friend in Silicon Valley said when I excitedly showed him those fresh-off-the-printer beauties. (man! It's hard to impress people in tech)
But it was a beginning, and I felt encouraged.
We drew inspiration for the components from Beepberry and Playdate. A tweet from Eric "The Pebble" Migicovsky set us on the right path.
I even have an email from Eric in my inbox, cheering me on. I plan to sell it as an NFT to my kids on my death bed. He doesn't know it, but he's sort of a hero to me.
The screen was the first thing to come to life.
The keyboard obediently followed.
It was coming together.
It was as if the universe was saying, “Yes, you can make this". Thanks to the internet, and a chance meeting with a Swedish hacker — I had a first version.
I proudly slammed these two beauties in front of my kids and said "You are the only kids in the world who have this". Their faces lit up. I was a cool dad. Life was good.
We had a brainstorming session to name the gadget, and my kids came up with Robodonkey. Each time they said "Robodonkey", they giggled. That was enough of a focus group for me.
If you’re thinking about building something, remember: even if you don’t know jack about hardware, sometimes all it takes is a little bit of madness and a lot of googling, copy-pasting code from chatGPT into Arduino IDE, and a Swedish engineer who is incredibly helpful (and likely needed money .. jk! jk!).
I did everything I set out to do — 1) made a gadget for my kids, 2) write a post and end it with an inspirational message.
Mission accomplished?
No! Not even close.
PART TWO
My kids with nothing held back told me the thing was unusable. It was tough to hear.
The outer shell of the device fell apart. The hinge was unstable. It was hard to type on the keyboard. The screen kept flickering.
It took one full minute to type "why is the sky blue".
Clearly, this needed fixing.
I reached out to 10 or so people who I thought could help. Depending on their backgrounds — mechanical, electronics, software — I would pitch the relevant problem. But I wasn't getting anywhere. It was the "wilderness period" for this project. Trough of sorrow, as they say. I met industrial designers, hardware hackers, 'maker space' gurus, even 'rapid prototyping' consultants — but I wasn't moving forward.
If I am being perfectly honest, I was mainly looking for shortcuts. I wanted someone who would just do this for me so that I could be the "idea guy". I kept going in circles.
I decided to dive into coding myself. I spent my days buying components and assembling them by following YouTube tutorials. While I wasn’t exactly a coding savant, it was infinitely more enjoyable than endlessly searching for collaborators.
Just when things were getting hard for me, one of my earlier contacts randomly emailed me and suggested I speak to someone named On. So I did, and he changed everything. It's because of him (and another brilliant soul I met soon after) that this device is now ready to ship.
On loved this project in the first 30 seconds of our chat. He thought a gadget for kids to ask questions was a fantastic idea. Or maybe he just felt sorry for me — hard to know. Either way, within days, we were working together day and night building the next version of Robodonkey. We were both putting in 10-15 hour days non-stop. It was an insanely fun time.
On told me to get rid of all moving parts. The hinge was giving me big headaches. In retrospect it is obvious, but I had been too fixated on the original design. Too in love with my original idea.
On was right. We needed a new design. We sent sketches and ideas back and forth — digital dictionaries, old phones, old computers, old apple devices, vintage handheld consoles — while we were marinating in nostalgia, one day On sent me this image.
Looking back, this was the moment things changed. No matter how many wild concepts we explored (at one point we got obsessed with "no more rectangles"), we kept coming back to this. It became our new direction.
I wish I had a collaborator like On in all my projects. He’s incredibly fast. Extremely open to feedback. Processes new ideas quickly. Dedicated to solving small or big problems. Genuinely fun and optimisitc.
Soon, we had a new PCB and prototype. We were buzzing with excitement. I have to admit, it was hard to sleep during those days. We were determined to make this thing real. And we did.
When I held the new version for the first time, it just felt right. The new design clicked. It was light, easy to hold, and best of all, no hinge problems—because there was no hinge. (there was a long and interesting period of experimentation with the keyboard. The keyboard deserves its own story some day)
The most important thing was that my kids were back to asking the silliest questions.
Who is the laziest president
What color is a donkey’s poop
I started showing Robodonkey to my friends and gave out about ten devices to parents with similarly aged kids. The feedback was incredibly useful. On and I addressed every issue. Everything kept getting better. The keyboard was nearly perfect (if I may say so myself).
My kids were what I'd would call in my previous job "Daily Active Users". It was clear we had built something easy to use, engaging, and appealing — a personal gadget for kids to ask their endless questions. It felt great.
On and I felt that we should try to sell this device. However, there was still one problem. The software kept crashing. It was really embarrasing every time it happened.
My friends would send me pictuers like this one. Both On and I knew that we needed a software specialist since my skills had run their course.
PART THREE
By May 2024, the hardware was looking fantastic. But the software? It was frustating. Key presses took a long time to register, crashes happened frequently, bugs were quite common. We knew this was not something we could hand over to kids. The code was a frankenstein mashup written by 3 people -- my original swedish friend, myself and chatGPT.
I had met many software folks and asked for help, but I couldn’t find the right collaborator. One night, at 2 AM, I called an old friend, Sid. We had worked together on some wild projects at my previous company. After that, he’d vanished. When we reconnected I learnt that he is learning Thai kickboxing. Ok, what!?
Sid’s life was a stark contrast to mine. He was like a wild mustang roaming the open wilderness, while On and I were more like domesticated workhorses, saddled with family life. But Sid called back and heard me out. He simply said, "Sure. Let’s do it." He wanted to re-enter polite society and liked the "goodness" of the project.
In the next few weeks, Sid rewrote everything from scratch. The keyboard became super snappy. The device hasn't crashed in weeks (and we have tried!). Sid made the software experience smooth. We simplified onboarding. It is now quite easy to get Robodonkey up and running.
Meanwhile, On had made stunnig progress on refining the design in many subtle but crucial ways. Each improvement made Robodonkey feel more like a "product" and less like a "prototype."
So, people of the world — here it is, a polished version that you can order for $100 and give to your kids. This is something On, Sid and I are incredibly proud of.
If you think this might be a cool gadget for your kids or some kid you know — please order it here. We would love to ship you one!