No One Would Hire Me, So I Created My Own Job

My First Time Working on a Startup

The first time I started looking for a job was somewhere around 2022, as a React frontend developer. I managed to go through a few interviews and realized that I simply didn't know enough yet. But even back then, for some reason, requirements for candidates were already starting to grow. Once I at least roughly understood what direction I had to move in, I started thinking about what to do next.

Now I kinda regret that when I was just starting out, I underestimated how useful it is to just build some boring project from scratch. Purely to practice a new technology or just to learn something. A todo app, a blog, something super basic. I started with a blog as my first proper pet project. Before that I had some small scripts and Telegram bots in Python, but nothing more serious. While learning different technologies, I also got into the whole self-hosted philosophy. I actually liked it a lot, especially because I had an old PC just sitting at home doing nothing. After playing around with it for a while, it turned into a pretty decent server. The only real downside was having no backup power during blackouts. I spent roughly a year messing around with the blog, hosting it on my own configured server, trying different technologies. But I still couldn't get hired.

During my last year of college, we got a really strong prof who had worked on his own projects before and had proper experience in IT. He also ran classes a bit differently. Instead of just doing some lab assignments from a manual, we were literally building our own projects during classes. We would come up with ideas, discuss with him whether they made any sense at all, whether it was possible to get traffic, and how we could test that. We made a few projects like that. A website with random tools for basically anything, we even tried building a classifieds board. None of it really took off and eventually all of it died. The teams were pretty big, everyone was responsible for their own part, and once some people started dropping out, keeping the project alive became super hard. You had to figure out someone else's code, deal with existing legacy, and nobody really saw much potential in those projects anymore. So they were just slowly forgotten. But if it wasn't for that prof, I honestly don't even know what I would have been doing back then. He was basically the first person who got me to try building not just some random thing, but a project that could theoretically be useful to someone.

After a few attempts like that, we once again had no project and started thinking about what to build next. I asked him if maybe he had some idea of his own that we could try to implement. That's how a classic AI startup for tracking sales team performance appeared. The idea was to pull in or upload call recordings, analyze them, and attach the result to a specific lead in a CRM. We built a basic version that more or less already did that. The team had two developers, a designer, and the prof himself as the product owner. At first I was just the frontend developer, later I started trying to lead the team...

By that point it was already the end of summer, and I also had to focus on getting into university. After I got in and spent literally one or two weeks there, I realized it was kinda shit. If I was going to study in Lviv, the only place I personally saw any point in was UCU. Polytechnic was pretty bleak. I mainly went there to meet people and hopefully find some like-minded people who were into the same stuff as me. Didn't really work out. Around that time I started carrying around the idea for Ramle in my head.

SolderDemon

After BeQuality got frozen and I managed to finish the semester at uni pretty quickly, I suddenly had a lot of free time again and the same question: what do I do next? Nobody really wanted to hire me either. Every time I got the urge to try the corporate world again, I would redo my CV, apply to relevant positions, and basically get nothing back.

A bit earlier, somewhere around August, I was told about a hardware project called rosco_m68k. It's a single-board computer built around an old Motorola 68010. The first time I calculated how much one unit would cost, honestly, I didn't really want to build it. But then a few people joined in, the purchase stopped looking that expensive and pointless, so I decided to try it purely for learning. Before that I had barely soldered anything and my hardware knowledge was pretty weak, but I quickly realized it was a really good way to learn how hardware and low-level programming actually work in practice. In my head I already had plans that the second the computer booted, I'd start hacking together my own OS in assembly, writing drivers and all that. In reality, we spent a few months just trying to get the thing to boot. Wrong component here, something else not working there, then we were screwing around with the firmware. It was brutal.

The idea originally came from my uncle, and he was also basically the one helping me debug everything because at the start I had no clue what was going on. While we were messing around with the board, I learned a ton of stuff in practice and generally got pretty hooked on hardware. Somewhere around that point I started thinking: maybe this could be something more than just one computer I built for myself.

Back in October, following all the classic lean startup rules, I made a simple landing page with a waitlist and started posting it on Reddit and in different Telegram chats. I honestly hadn't seen a more toxic audience than the people in Polytechnic chats at that point. But literally a few hours later I started getting email notifications that someone had actually left their email on the waitlist. For me that was a pretty strong signal that I wasn't the only one interested in this stuff, and that SolderDemon was probably worth pushing further.

Once I had more or less figured out how to assemble the computers properly and saw that there was an audience, I started building my own e-commerce site. As simple as possible, no admin panel and no pile of unnecessary features. Around that time Claude and Codex agents got a lot better, and SolderDemon became the first project where I properly started experimenting with vibe coding. In a pretty short amount of time I already had a store with products, a cart, and basically everything I needed to launch.

The biggest problem left was payments. Stripe wasn't really an option for me directly at the time, so I had to try a bunch of alternatives. At some point I realized it would actually be easier to register a company in the United Kingdom and connect Stripe through that. So that's what I did. I added payments, finished filling the store, and then at the beginning of January I got a notification that money had come in. SOMEONE BOUGHT A KIT. I can't really describe that moment properly. I don't think I'd felt that kind of high from a project before.

After that, SolderDemon kept going pretty well. I figured out international shipping, added new stuff, and slowly learned a bunch of things I previously had absolutely no reason to learn. Whenever I hit an idea crisis and had no clue what to do with the project next, I would just switch to something else. Then after some time, new ideas for SolderDemon would appear, and it basically kept going in circles like that. During one of those idea crises, I finally decided to try launching Ramle.

Ramle

Ramle happened pretty spontaneously. I looked at the trend around finding friends, saw that there was actually decent search demand for it, and thought, why not. To make it easier to actually ship, I asked a friend to build it with me. He took the backend, I took the frontend. We worked on it for about a month, and after another big update it became obvious that there was no point sitting there polishing it without any users. We had to release it. Of course there were bugs, mistakes, onboarding wasn't thought through that well, but at that point real users had to show us what was actually wrong.

For the first two weeks in production, basically nothing happened. We kept polishing things, while I tried to promote the project somehow. Posted on X, Threads, Product Hunt. Then pretty unexpectedly, different social media pages started writing about Ramle and real users finally started coming in. Together with them came problems we hadn't even thought about. What surprised me was how many people actually messaged us themselves about bugs or things they didn't like, and we tried to fix all of it as fast as possible. I already told the full Ramle story in more detail here, so I won't repeat too much here.

The project lived for around three months, but those three months were super intense. I worked a lot during that period. At some point we hit the main problem with any social product like this: we somehow needed to reach critical mass. Paid ads were expensive, and we never really found a way to consistently bring in enough people through free marketing. In the end, we decided to shut Ramle down.

And that was kind of a weird moment. I still couldn't really get hired, but by that point I had already managed to build and launch a product with a team, get real users, deal with production bugs, look at analytics, try marketing, and eventually shut the project down when the economics just didn't work.

What's Next?

And next? I don't really know. I tried sending out my CV again. Mostly silence, or at best, "we decided to move forward with another candidate." I think I finally have some idea of how to keep growing SolderDemon, and I'm thinking more and more about focusing on hardware. At the very least, I'm genuinely interested in it, I can make a bit of money from it, and in the long run it's also harder to automate with agents.

I've seen a lot of people saying there aren't enough miltech and hardware specialists. But if you actually look at the job listings, what they usually mean is that there aren't enough seniors. Though to be fair, there aren't enough seniors anywhere. The only question is: who is supposed to grow those seniors if nobody wants to hire juniors?