In August 2025, four high school students from Mundelein, Illinois decided to build a hardware startup. None of us had done it before. We had no investors, no advisors, no engineering degrees. What we had was a problem we understood deeply, a group chat that never went quiet, and the kind of recklessness that comes from not fully understanding how hard something will be.
Seven months later, Nexus Pen is a real product with a real website, a working firmware stack, a mobile app on the App Store, and pre-orders open at $119. This post is the honest story of how we got here — the wins, the failures, and what we'd tell other young founders who are thinking about starting something.
The Founding Story
The idea came from a specific moment of frustration in AP Chemistry. I was taking notes during a lecture on reaction mechanisms, hit a concept I didn't understand, and reached for my phone to look it up. Twenty minutes later I was watching YouTube videos that had nothing to do with chemistry. The distraction wasn't the goal — it was just what happened when I opened a general-purpose screen.
I started thinking: what if there was a way to get the answer without opening the portal to everything else? The pen was already in my hand. What if the pen could answer the question?
I brought the idea to Chase, Michael, and Izic — my closest friends and, it turned out, my cofounders. We spent three weeks arguing about whether it was a real idea or a stupid one. By the end of those three weeks, we had a working proof of concept: an ESP32 development board, a small OLED screen, and a rudimentary Bluetooth connection to a Python script that called the OpenAI API. It was ugly, but it worked. That was enough.
Learning to Code (for Real)
I'd been tinkering with code since middle school — mostly Arduino projects, some web stuff. But building a real product required a different level. The firmware alone involved C++ on ESP32, FreeRTOS task management, BLE GATT protocol implementation, I2S audio drivers, and memory management on a microcontroller with 320KB of RAM. None of that was covered in AP Computer Science.
The learning process was brutal and nonlinear. I'd spend three days on a bug that turned out to be a single missing semicolon in a preprocessor directive. I rewrote the BLE audio pipeline four times before getting latency low enough to be usable. The audio buffer size caused an entire class of firmware crashes that took two weeks to diagnose — a DRAM overflow that only manifested at runtime, not at compile time.
What I learned is that you don't actually learn to code by taking courses. You learn to code by having a specific thing you need to make work, failing repeatedly, and eventually understanding the system well enough to fix the failure. The frustration is the education.
The First Prototype: Honest Assessment
Our first real prototype — assembled in my bedroom in September 2025 — was terrible in the ways that matter and incredible in the ways that count. It was terrible physically: a 3D-printed shell held together with hot glue, components wired point-to-point on a breadboard, battery life of about 45 minutes. It looked like a prop from a student science fair.
But it was incredible functionally: you pressed a button, asked a question, and heard a clear answer through a pen-sized speaker 2-3 seconds later. The OLED showed the key text. It was distraction-free by definition because there was nothing else on the device. Every person who tried it had the same reaction — genuine surprise that it actually worked.
That moment — watching someone pick up the pen, ask a question, and say "wait, that's actually useful" — was the validation we needed to keep going. Ship ugly, learn fast.
The Mistakes That Cost Us the Most Time
We made a lot of mistakes. These three cost us the most time and taught us the most.
Building before validating. We spent two months building features that users didn't care about — specifically, a complex note-taking mode that stored handwritten notes on the device. When we showed it to potential customers, nobody asked for it. The questions they asked were all about the AI response quality and the speaker volume. We'd optimized for the wrong things. Talk to users before you build, not after.
Underestimating the supply chain. Hardware startups live and die by supply chains. We learned this when our original OLED supplier went on backorder for six weeks with no warning, halting a prototype run. We now qualify at least two suppliers for every critical component and keep 20% buffer stock. Build redundancy into your supply chain from the beginning — never rely on a single source for anything that would stop you cold.
Avoiding the hard conversations. Early on, we avoided having direct conversations about equity, roles, and compensation because those conversations felt uncomfortable. That avoidance created ambiguity that caused real tension later. We eventually had all the hard conversations — and they were fine. The anticipation was worse than the reality. Have the uncomfortable conversations early, when the stakes are low.
What High School Actually Gives You
People sometimes treat being a high school founder as a disadvantage — no experience, no network, no credibility. But there are real advantages that don't get discussed enough.
We are the target market. We sit in classrooms every day. We know exactly what the problem feels like from the inside, and we know immediately when a solution is heading in the wrong direction. That product intuition is genuinely rare and genuinely valuable.
We have low overhead. None of us has rent, a salary requirement, or investors asking about burn rate. That freedom lets us make decisions based on what's right for the product rather than what's right for this quarter's numbers. It's a form of optionality that disappears quickly in adult life.
We have nothing to lose. This sounds trite but it's actually profound. We're not risking a career, a mortgage, or a retirement account. The worst case is that we learn a lot and the company doesn't work out. That risk profile allows a kind of boldness that gets harder to maintain as the stakes get higher.
Advice for Young Founders
If you're in high school and thinking about starting something, here's what I'd actually tell you.
Start with a problem you personally live with. The best startup ideas come from genuine frustration with a problem you understand from the inside. We built Nexus Pen because we needed it. That kept us going through months of setbacks in a way that a business opportunity we'd read about in a blog post never would have.
Build something that works, even if it's ugly. Talking about your idea is not the same as building your idea. Get something working as fast as possible, even if it's embarrassing. The working prototype earns you credibility with cofounders, early users, and eventually investors. The pitch deck does not.
Find people who are better than you at the things you're worst at. I'm strongest on the technical side. Chase is strongest on operations and business. Michael is strongest on finance. Izic is strongest on marketing. That diversity of skills is why we can execute across all the dimensions a hardware startup requires simultaneously. Find your Chase, Michael, and Izic.
Don't wait for permission. No one will give you permission to start a company. No one gave us permission. At some point you just decide you're doing it, and then you do it. That decision is the only prerequisite.
Where We Are Now
Seven months in, Nexus Pen is real. The firmware runs reliably. The app is on iOS and Android. The website is live. Pre-orders are open. We're a long way from where we started — and honestly, a long way from where we want to get to.
But we built something from nothing, in high school, with no outside help, and it works. Whatever happens next, that's true. Nobody can take it back.
If you're a young founder thinking about starting something, I hope this helps. The path is harder than you expect and more worth it than you can imagine. Build the thing.