The things that wouldn't exist
On September 4th I shipped version 1.0.0 of “Rudra Drum Learning” to the Apple and Google stores. It’s an important milestone for me, the first commercial product of my new company, Aeris Nova. I’ve worked as a contractor for many years, but this year I ended up building my own product. It wasn’t a premeditated decision, it just kinda happened.
My son started learning to play the drums. His teacher is a Cuban guy, so we’re both Hispanics in a French speaking country. Speaking your own language is nice when you live abroad, and it helps build rapport. As the weeks go by and I take my kid to class, one day the teacher invites me to try too, and so I also start learning the drums.
In class the teacher used a few apps, and one that worked well was based on the rhythm circle, or circular metronome as it’s also known. Basically a ring of circles where you draw loops of notes. This simple idea makes rhythms easy to visualize and lets you represent any combination. In the circle you’re not limited to traditional structures like 3/4 or 4/4. Every instrument can loop with its own rhythm. So you can write very easy patterns, or you can go wild and do very complex polyrhythms.
Eventually we decided to buy a drum kit for home. We went for an electronic one, because a regular kit is super loud, and well, we have neighbors 😉. When the e-kit arrived I started looking for practice apps and subscribed to Melodics. Their product is very good, but it’s only in English. What a silly detail, right? But it isn’t. I care. Maybe it’s because I use a different language at home, in town and at work, and I know learning in your own language just feels different. Sure, I speak English and can explain things to my kids, but I feel they should be able to learn in the one they prefer.
So I kept thinking about that rough circle web app and its limitations, and I set out to make my own, for my son’s teacher in class and for us as students. How hard could it be? With Claude it’ll probably be a piece of cake, right?
Claude Opus 4.5 was the hotness when I started. I was drunk with power, and after a few nights I had a first version working. I deployed it to Cloudflare Pages and we had something ready for class. Cool! The teacher connects his iPad to a speaker and a projector, and that’s it. Over the next couple of weeks he’d find bugs and I’d fix them, add features, polish, etc.
But I wanted something that could give feedback to the student, because when you practice at home you need to know if you’re doing it right. A couple more nights with Claude figuring things out, and I had a working version with Web MIDI. Nothing super fancy, but it worked. Ok, but now I wanted it on the iPad, and Safari doesn’t support Web MIDI. Ok, I’ll build an app. But I’ve never built an app. I don’t know Swift, Objective-C or Kotlin, and Java is a bad memory from a distant past.
Before AI, that’s where the dream would have stopped. The web part I could have done. Without Claude I don’t know if I’d have put in the time for that first web version, but at least I had the skills. A mobile app is a different beast. I’d done a few iOS tutorials in the past, and that’s it.
So, still under the effects of a Claude-induced delusion of massive speed gains and limitless possibilities, I decided to embark on this mission and build a native app. Not being totally reckless, I picked a stack that could work for me: Tauri. I could reuse a big part of the frontend and get multi-platform support for free. It seemed sensible: I could go into uncharted territory but stay close to what I know. And again, after several nights over a few weeks, I had the first working version.
But Apple doesn’t want you sideloading apps, so after a week you have to reinstall them. Ok, let’s see how to get this done. Of course you have to pay. You need an Apple developer account and a yearly $99 fee. Fine, I’ll pay it. But if I’m paying anyway, maybe I can sell this app too. Or so I thought. Boy, was I naive. I have to register the company and get a DUNS number. Ok, fine. It takes a few days and I get the number. Now the bank account, the tax forms, etc. Ok, fine. So after I deal with the bureaucracy I can start building the app in test mode. Ok, this is progress.
Now I have to learn about subscriptions, and how to set up the store listings, and I’ll need marketing material, screenshots, etc, etc. And don’t forget I’m doing all this for Google’s Play Store too.
At this point I’ve invested quite a bit of time in the project and I want to take it to the finish line. And I’m thinking, if I’m going to charge money for it, it needs to be nice and polished. And it should work on the iPhone too. Designing for such a small screen is a challenge, you don’t have as much screen real estate as on a computer or an iPad, so I redesign everything. And the polish takes so long. Claude doesn’t feel the app, I need to feel it. Every round shows more details and edge cases, and the work grows and grows.
So I decide to draw the line and focus on shipping version 1.0.0. I spend the next weeks on QA, only bug fixes and polish. No more new features, no more redesigns. It took me a bit over 150 hours to get to this point. That’s a month of work on top of my job, my family, etc. But it’s done.
This was so much more than I expected. Releasing a commercial app and trying to sell it has so much friction. There’s so much to do and learn, and it’s not the technical part, it’s everything around it. No wonder companies have full teams for this. And yet here I am, on my own. I had no idea how much I’d have to learn, and I have no idea what’s still to come, because now I actually have to sell and market this.
But here we are. A new school year has just started. Soon music classes begin again, and all the students at the teacher’s academy will start using the app. They’ll be able to practice at home, in their own language. We’ll see where this adventure takes us.
If you made it to the end of this massive brick, you’re most likely a friend. If not, well, you’re my hero XD. And don’t forget to check out Rudra!
Cheers!