2025 SoloDevStack Wrapped: A Year of Building, Learning, and Getting Hacked

2025 is wrapping up, and what a year it's been.

I learned new languages. I shipped real projects. I got my server hacked. I had my first developer interviews. I started building things I actually want to use. I experimented with AI, game dev, and about a dozen other things that went nowhere.

This is my 2025 wrapped. The wins, the failures, the lessons, and what's coming in 2026.

New Technologies I Actually Learned

Astro: Finally, a Framework That Makes Sense

This blog runs on Astro. And honestly? I love it.

I've used React, Vue, Next.js, and a bunch of other frameworks. They're fine. But for content-focused sites like blogs, they always felt like overkill. Why am I shipping 300KB of JavaScript for a page that's mostly text?

Astro gets it. Ship HTML by default. Add JavaScript only where you need it. The "islands architecture" thing actually works. My blog is fast, simple, and I understand every part of it.

What I like about Astro..

• Zero JavaScript by default. Pages are static HTML unless you explicitly add interactivity.
• Component flexibility. I can use React, Vue, or Svelte components if I want, or just write plain Astro components.
• Content collections. Built-in support for markdown/MDX content with type safety.
• Fast builds. The dev server starts instantly. Builds are quick.

It does exactly what it's supposed to do. No more, no less. I'm a fan.

Go: Learning the Hard Way (On Purpose)

I started learning Go towards the end of the year, and I'm doing it the old-fashioned way. No AI assistance. No Copilot. Just me, the documentation, and Stack Overflow.

I wrote about this in my Rembiti post. The short version is that I want to actually understand Go, not just accept AI suggestions and hope they work.

My impressions so far..

Go is simple in the best way. There's usually one obvious way to do things. The error handling is verbose but makes you think about failure cases. The tooling is excellent. And it's fast enough that I don't have to think about performance.

I'm building Rembiti's backend in Go as my learning project. It's slow going, but I'm actually retaining knowledge instead of just copy-pasting.

By summer 2026, I want to be genuinely competent in Go. Comfortable enough to use it in interviews, build microservices, and handle devops-style work. That's the goal.

Rust: A Brief Affair

I also dabbled in Rust this year because I was building Life OS with Tauri.

Tauri is great. It lets you build desktop apps with web tech but without Electron's bloat. The catch is that the backend is Rust.

I learned enough Rust to be dangerous. Basic syntax, ownership concepts, how to fight with the borrow checker. But honestly? I'm thinking of switching Life OS to Wails and Go instead.

Here's my reasoning..

If I need to learn a systems language, I'd rather focus on Go. It's more relevant to the backend/devops work I want to do. Learning both Go and Rust simultaneously is spreading myself too thin. Wails gives me the same "desktop app with web frontend" approach, but with Go on the backend.

So Rust might be a 2025 experiment that doesn't continue into 2026. We'll see.

Projects I Actually Worked On

Apatero: The Wild Ride

The biggest project of my year was Apatero. An AI content generation platform. It's live now, and getting there was an adventure.

What I learned building Apatero..

Docker and real deployment. I'd done local development and basic hosting before, but Apatero required actual infrastructure. Docker containers, proper server setup, managing multiple services, SSL certificates. The stuff that tutorials gloss over.

Security, the hard way. Yeah, I got hacked. Someone compromised my server. It was a wake-up call. I learned about proper firewall configuration, SSH hardening, monitoring, and all the security basics I'd been ignoring. Getting hacked sucked, but it made me a better developer.

Persistence. Apatero isn't finished. There's still a ton of work to do. But at least it's live. At least people can use it. That's more than most projects ever achieve.

Current status..

We're live at apatero.com. Still figuring some things out. Still adding features. But the foundation is solid, and hopefully Q1 2026 we start actually earning revenue. That's the goal.

Life OS: Privacy-First Everything

I wrote a whole post about Life OS. It's a personal productivity system that runs entirely locally. Journal, tasks, calendar, contacts, with a local AI that can search your data.

This is the project I'm most excited about because I actually need it. I want a single app for organizing my life that doesn't send my data to someone else's servers.

The tech decision I'm wrestling with..

Currently built with Tauri (Rust backend). But I'm seriously considering switching to Wails (Go backend). The functionality would be the same, but I'd get to use Go instead of fighting with Rust.

I haven't made the final call yet. Either way, Life OS is continuing into 2026.

Rembiti: Birthday Reminders Done Right

Rembiti is a birthday reminder app with contact notes and message templates. Simple concept, but nothing out there does it the way I want.

This is also my Go learning project. Building the backend without AI assistance, understanding every line of code.

Current status..

Mock UI phase. The designs are done. Backend is in progress. This one has a longer timeline because I'm deliberately slowing down to learn properly.

Travel Tracker: For Couples Who Explore

The travel tracker started because my girlfriend and I couldn't remember all the places we'd visited together.

It tracks countries AND cities, is designed for couples to share, and includes notes and memories for each place.

Current status..

Architecture in place, basic flows working. Not ready to show yet, but progress is happening.

The Linktree Clone (Unpublished)

I also worked on a Linktree-style app with a friend. Co-owned project, built with Django and React because that's my comfort zone.

We didn't publish it. Life got busy, priorities shifted. But it was still a good learning experience. Working with someone else, even on a side project, teaches you things solo work doesn't.

Not everything has to ship to be valuable.

First Developer Interviews

This was a big milestone for me.

I had my first real developer interviews this year. Not just casual conversations, but actual technical interviews with coding challenges and system design questions.

The first interview..

Opened my eyes to how much I still need to learn. I wasn't terrible, but I wasn't great either. There were gaps in my knowledge that became painfully obvious. It was humbling.

The second interview..

Made me realize I'm not as bad as I thought. Did two technical rounds. First one was decent. Second one I felt genuinely good about.

I'm still waiting for the final response. Should hear back in January 2026. Either way, the experience was valuable.

What I learned..

I need to know more. That's not self-deprecation, just reality. By summer 2026, I want to be confident in interviews. Especially for backend roles, Go, microservices, devops-adjacent work.

The way to get there is just more building, more learning, more practice. No shortcuts.

Random Experiments That Went Nowhere (Or Somewhere)

2025 wasn't just focused projects. I experimented with a bunch of stuff.

AI Agents and Chatbots

Built some AI agent prototypes. Played with chatbot implementations. Nothing shipped, but I learned about prompt engineering, context management, and the limitations of current models.

The AI space moves so fast that anything I built would be outdated in months anyway. But the understanding transfers.

Training an LLM

Tried training my own language model. Didn't go great. Turns out you need serious compute and serious knowledge to do this properly. I had neither.

It was a for-fun experiment, not a serious attempt. I learned enough to know I don't want to be an ML engineer.

Training LoRAs

This one actually went somewhere. I trained a ton of LoRAs for content generation, mostly for clients. This ties into the Apatero work.

LoRA training is more accessible than full model training. The feedback loop is faster. You can actually see results without a GPU cluster.

Game Dev with Godot

I bought a course. I bought assets. I had visions of a farming game. A little cozy experience where you grow crops and build a community.

Then reality hit. I don't have time for game dev right now. Not with all the other projects I'm juggling.

But I see you, farming game. I envision you. Maybe 2026, maybe later. The assets are waiting.

Opinions I've Formed This Year

A year of building gives you opinions. Here are mine.

On frameworks..

Use boring technology until you have a specific reason not to. Django, Rails, PostgreSQL, vanilla JavaScript. They're not exciting, but they work. Save the cutting-edge stuff for when you understand why you need it.

On languages..

Go and Astro both embody "do one thing well." That philosophy resonates with me. I'd rather use tools that are focused and opinionated than tools that try to do everything.

On AI assistance..

It's incredible for productivity. I use it constantly for work. But for learning? Sometimes you need to struggle. The AI can't give you understanding. Only frustration and eventual breakthrough can.

On shipping..

Done is better than perfect. Apatero isn't finished, but it's live. That matters more than having every feature polished. Ship, learn, iterate.

On security..

Take it seriously from the start. Getting hacked was a painful lesson. Learn firewall basics, SSH hardening, monitoring, before you need them.

What's Coming in 2026

2026 is going to be big. Here's what I'm planning.

Q1: Apatero monetization. The platform is live. Now it needs to make money. That's the priority.

Q1-Q2: Life OS completion. Whether I stick with Tauri or switch to Wails, I want Life OS functional and usable by mid-year.

Throughout the year: Go mastery. Continue building Rembiti. Maybe pick up more Go projects. By summer, I want to be interview-ready for Go roles.

Ongoing: Mobile apps. Rembiti and the Travel Tracker are both mobile projects. At least one of them ships in 2026.

Dream goal: Backend/devops role. By end of 2026, I want to be confidently applying for backend positions. Go, microservices, infrastructure. That's the direction I'm headed.

Secret goal: That farming game. Probably not 2026. But a guy can dream.

Lessons From 2025

Slow progress is still progress.

I wrote about this before. Some weeks I barely touched my projects. But I kept coming back. Small consistent effort beats sporadic marathons.

Getting hacked teaches you more than any tutorial.

Nothing motivates learning security like watching someone compromise your server. Painful but effective.

Interviews reveal gaps.

You don't know what you don't know until someone asks you about it. Even failed interviews are valuable feedback.

Building for yourself is motivating.

Every project I'm excited about solves my own problem. Life OS because I need it. Rembiti because I forget birthdays. The travel tracker because my girlfriend and I wanted it.

You can learn in public.

This blog, these project updates, sharing the journey. It's scary but it creates accountability. And maybe helps someone else who's figuring things out.

Final Thoughts

2025 was a year of building foundations.

I didn't get rich. I didn't go viral. I didn't have some massive breakthrough that changed everything. But I learned technologies that will serve me for years. I shipped a real product. I started projects I genuinely care about. I had my first interviews and discovered I'm not as far behind as I feared.

That's progress.

2026 is about execution. The foundations are laid. The projects are in motion. Now it's about finishing what I started and building the career I want.

Let's see what happens.

If you're on your own solo dev journey, feel free to reach out. Share what you're building, what you're learning, what's working and what's not. We're all figuring this out together.

Here's to 2026.

FAQ

What was your biggest win of 2025?

Launching Apatero. Getting a real product live, on real infrastructure, with real users. That's the milestone that mattered most.

What was your biggest failure?

Getting my server hacked. It was embarrassing and stressful. But it forced me to learn security properly, so maybe it was a win in disguise.

What technology are you most excited about for 2026?

Go. I'm committed to getting genuinely good at it. The simplicity appeals to me, and it's relevant to the backend/devops work I want to do.

Will you ever finish that farming game?

Probably. Eventually. The vision is there. The time isn't. Maybe late 2026 or 2027.

Any advice for developers starting their solo journey?

Build things you'd actually use. Share your progress publicly. Don't wait for perfection. Get hacked early so you learn security before it really matters. And remember that slow progress is still progress.