Rossi Stefano

Since young age I played with computers. One day I discovered that I could program them to do things for me. I started doing hello worlds in C++, and some year later I was coding small 2D games in java applet for fun, like:

One day my friends started playing sudoku. So I used my programming skills to developed a sudoku solver with rule's checking and recursivity. That's how I won my sudoku's supremacy.

One day I was reading about the Manhattan algorithm and realized I could implement an adaptive intelligence for the ghosts of my pac man clone, so I did it. This was my first rudimentary approach to AI.

Years passed and got some education and working experience in IT developing and deploying control-command solutions for automated tunnel systems. But I kept hearing about AI, especially with the advent of large scale Natural Language Model. So I started a bachelor in Data Engineering in Fribourg: Here an example of how AI and Ethics can be used in real-world applications.

Swiss law is sensitive to privacy and security in IT and complexity is not enough to prevent dangers. Here you can find an example of Backdoors in LLMs.

I also manage an enterprise ready kubernetes cluster with support for cloud native applications, high availability, dynamic scaling, backups, metrics, and blue-green gitops. In this cluster I run my personal projects and other services, HA databases, some websites, and more. This allows me to have full governance over my data and the software I use, while also providing a platform for testing, developing and deploying. I am a Linux enthusiast, using windows as complement.