From school assignments to object detection with YOLO

Scroll all the way down my GitHub and the first thing you’ll find isn’t AI at all. It’s Lector, a program that compared texts (March 2024), and its sibling Lector.pdf, which did the same with PDFs. Around the same time there’s Maquina-de-cafe, a coffee machine. That’s how almost everyone starts, and that’s fine.

That repo also has about ten pull requests from the same day called “Create uno.txt” or “Update uno.txt”. That was me learning git. Nobody is born knowing how to merge.

The Java and C# era

In 2025 school kept me busy with Java and C#: Practicas_escolares_Csharp, Practicas_java, my first steps with JavaFX, a Laberinto (maze) in C# and a repo I very confidently named java_Avanzado. None of it will change the world, but that’s where I learned the boring stuff that actually matters: object-oriented design, structuring a project, and not giving up when something won’t compile.

The first project that felt like “AI”

In October 2025 I built YOLO Vision: real-time object detection with YOLO and OpenCV. It was the first time I saw a computer “understand” something, even if it was just drawing a box around a mug and saying cup 0.87. That’s when I got hooked.

Then came an SSL-encrypted chat, a Pygame game with agents that patrolled and chased you, and an online Pokémon battle game. Still “learning projects”, but each one closer to what I wanted to do.

What I take from it

I didn’t delete any of it. I like that the path shows: from comparing two .txt files to, a few months later, running a RAG system in production. If you’re just starting and feel embarrassed by your GitHub, don’t hide it. It’s your history, not your portfolio.