Year Three Review

The necessary detour

After my dream project crashed, as I wrote about in last year’s review, I stayed 50% with my old client. The plan was to spend the remaining 50% building my own apps, finally creating something of my own.

Except I couldn’t. My mind wasn’t there.

I also got some pushback for staying. People questioned my choice, my commitment to freelancing, and even my courage. My closest colleagues always supported me, but hearing those comments and seeing those smiles hit me harder than they should have, especially since I was already down.

I thought burnout would be struggling to get out of bed, but for me it was looking at code, something I’ve loved for decades, and feeling physically ill. So even though the client work started well enough – building an accessible design system together with great people – it still did nothing to help me get over my burnout.

I needed something completely different. So, instead of struggling with building apps I couldn’t focus on, I built something physical. A small cabin behind our house. Wood, glass, and actual progress. Every wall that went up was proof I could still create something. When my programming identity was in ruins, at least I built a house.

The slow revelation

Building that cabin did exactly what I hoped it would. By switching between physical construction and digital work, my brain started working again. Not on building apps or writing code, but on understanding what the hell had happened to me.

Three things became brutally clear:

1. The market is dead. Not for everyone, of course. But the recent wave of layoffs flooded the market with talented developers willing to accept ridiculously low rates just to keep working. It’s driven many of us who wanted to stay freelance back into employment – especially if you want lead roles. I wrote about this last year, but this year it got real.

2. AI will eat our jobs. Maybe not tomorrow, but way sooner than most think. If you think AI won’t fundamentally change what it means to be a developer, you haven’t been paying attention. The repetitive stuff, the boilerplate, and even the “implement this really well-specified feature” – AI is getting seriously good at that. We’ll still need developers, just not the same kind.

3. I might not be a developer anyway. This was the big one. After 23 years, I had to ask: Am I even a developer anymore? I still love code. I love checking out new CSS features, I love doing SwiftUI experiments, and exploring the latest frameworks and techniques. But writing code eight hours a day? Building that same validation logic for the tenth time? That’s simply not me anymore.

   I love designing systems, architecting solutions, leading teams, making the bigger picture decisions. Don’t get me wrong – there are still coding challenges that get me excited, but most programming work just isn’t that.

So there I had it: The market is dead for someone like me. AI is coming for my job. And I might not even be a developer anymore. These three revelations should have sent me into panic. Instead, they felt like permission. Permission to stop forcing myself into a role that no longer fits. Permission to admit what I actually want to do. Permission to change.

Finding my way back

Following that clarity, something shifted this summer. After months of avoiding code, I finally figured out how to make AI tools actually work for me. You have to be brutally specific about what you want and what you don’t want, specify your architecture, your design patterns, even your coding style guides.

Suddenly I wanted to build things again. Claude Code and others handled the boilerplate while I focused on the interesting problems. No more eight hours implementing form validation. Instead, I could spend that time making the decisions that matter. The productivity boost was amazing, but more importantly, when I realized how fast we could build things now, I started thinking bigger.

There was this product idea I’d been carrying around for years but never thought I could pull off. Unlike the small productivity apps I’d been working on – which I still think would be useful – this was something much bigger. I can see how it starts small and grows into something substantial. Multiple revenue streams. It can go in multiple directions. But it’s still too big for just me, even with AI making everything faster. I’ll need co-founders, and that will be fun.

For the first time in over a year, I felt like myself again.

The path forward

That breakthrough changed everything. Suddenly there are so many things I want to try with this product idea, so many people I want to talk to about it. I find myself thinking through features and business models during walks, jotting down ideas, and actually building it whenever I can steal a few minutes.

But here’s the thing – I still really like client work. I genuinely love working with great people at companies where I can make an impact. My product dream is still years away, and in the meantime I want to really lean into the work that energizes me most: Architecture.

I got a taste of this kind of work again after my summer vacation when I got pulled into a week-long architecture session with some of the best front-end developers I know. Five days of whiteboards, system diagrams, deep discussions about both technical solutions and the business problems we needed to solve. I loved every single minute. Not the “this is fine” kind of love. The “I could do this forever” kind – the same feeling I had during that dream project. This was the work I actually want to do.

So I’m making some commitments for this year:

• Study and write about architecture every two weeks, documenting the decisions and patterns I’m using to build my product along with practical tips for applying architecture in development flows
• Push for more architecture work within my current assignments, steering conversations toward the bigger picture whenever possible
• Get technical certifications like AWS Solutions Architect and learn more about the business side through enterprise architecture certifications like TOGAF
• Start submitting talks to conferences again

I’m hoping this leads to assignments that actually fit me better. The market is what it is, but maybe if I keep building expertise and visibility in architecture, the right opportunities will find me. And who knows – maybe the product will become something real along the way.

Final thoughts

Until then, year three gets a C+. Not for the billable hours – they barely covered my expenses. Not for traditional success. But for finding my way back up again and getting to a place where I’m actually excited to build stuff again.

Even with tight finances, part of being a freelancer is that I can decide to spend what I have on what matters to me. For me, our freelancer bootcamps, long vacations, and conferences are required, not optional. Getting to CSS Day in Amsterdam with Jessica, my closest freelancer friend, every year is the best – I still love all the new CSS techniques and web stuff, and having someone to process all the ideas with makes it so much better. Beyond Tellerrand was something completely new for me this year and really opened my perspective, but probably not where I’ll focus going forward.

Looking ahead, I need to start attending more architecture conferences. If I’m really going to lean into this direction, I need to be where those conversations are happening. Plus, all that business-side learning will help with the product idea that’s still driving my excitement.

Here’s to year four: The year of architecture

• article
• freelancing
• personal