This is a submission for the Gemma 4 Challenge: Write About Gemma 4

It started with a real need. Then came the idea. And then a vision — to turn this idea into a reality.

When I started my journey with dev.to, I never imagined it would lead me to building my own app. I'm a Technical Writer. My job is to write and manage documentation, not to build apps. I joined the platform to be more up-to-date with current technologies, connect with the community, and write some articles to describe the nuances of working as a tech writer. At the beginning, the DEV Challenges tab was a place I immediately labelled in my head as a no-go zone. I thought, "I'm not a developer, this is not for me." It's hard to believe how much has changed in a little more than one month.

The first ideas came to my mind when I saw posts related to the Earth Day challenge. Some of them I found really inspiring, and I thought: "Wow, these people are so talented! They can do so much with their coding skills. They can actually make other people's lives easier!" At that moment, the idea was born — "It would be super cool to have such skills to help Polish people with recycling." But it was still vague. Still in the area of dreams. I still thought it was a task for someone wiser and more skilled; definitely not me.

Challenge that changed everything

Then I saw the Gemma 4 challenge and something clicked. The puzzles fell into place. "What if I could make AI analyze photos of waste and tell people which bin to throw it in? What if I could build an app that does that? It could be useful for so many people — and a great way to learn how to build something myself." The idea was still scary, but it felt tangible. Like something I could actually try.

I opened a new chat in Claude and started writing. I wrote down the idea, the whole vision of how the app should work, the features I wanted to implement. I described the issues Polish people have with recycling, and how I wanted to solve them. I wrote down the questions I had, and the doubts and fears that were holding me back. And then I asked for help. "How can we make it happen? Where do I start?"

What and how I built

I consulted my idea with Claude, and then I started building. I followed the guidance I got, but I also made my own decisions, challenged Claude when I disagreed, and corrected it when its suggestions didn't match the business logic or user needs.

Could I have done it better as a developer? Of course. I started by running Gemma 4 locally with Ollama, and even though it was satisfying to see it work, my laptop froze under the load 🥶 Doing proper research and finding the right tools took me longer than it would take a professional. I was like a child in the dark – learning new tools, debugging, making mistakes along the way. But every time I solved a problem or implemented a feature, I felt a sense of accomplishment that I hadn't experienced in a long time.

While I was the brain behind the scenes, AI was the extension of my capabilities. Following my instructions, Claude created the boilerplate code in Next.js, adjusted styling in Tailwind CSS, and even created icons that I imagined and described. With all the necessary pieces in place, Gemma became the heart of the project — analyzing images and providing recycling guidance 💛

You can read about the app I created in this article:

What's in it for me

When was the last time you did something for the first time? For me, it was building this app. It was a rollercoaster of excitement, frustration, doubt, and finally pride when I saw the final product working. It made me realize I'm capable of more than I thought.

The Gemma 4 challenge broke a barrier in my head – the one that said "I can't do this, I'm not a developer, this is not for me." The most important lesson: if something sparks curiosity, why not try? You don't have to be an expert to start. You just need an idea, a vision, and the willingness to learn and experiment. The tools are there, and AI can be a powerful ally in turning your ideas into reality. You just need to take the first step.

Yes, I prompt-engineered the app. Yes, I used AI to do the heavy lifting. But I was the one who had the idea, who made the decisions, who guided the process, who put everything together. I was the one who turned a vague idea into a real product that people can use. And that's something I'm really proud of.

If you've ever told yourself "this isn't for me" about something you secretly wanted to try – I see you. Try it anyway. You might surprise yourself.