Ruby Is the Best Language for Building AI Web Apps

Carmine Paolino

I’m Carmine Paolino. I wrote my first scripts at 5 to automate launching my favorite games. I’ve been hooked ever since.

Right now I’m building RubyLLM, an open-source library that gives Ruby a clean, provider-independent interface to every major LLM. I also run Chat with Work, an AI tool that lets you talk to your actual work data. Before that I co-founded Freshflow: AI that tells supermarkets how much fresh produce to order so less of it ends up in the bin.

I got into open source early, running the Italian Gentoo Linux community as a teenager. Studied CS at the University of Bologna, where I published work on large-scale social network analysis with Springer. Then AI at the Free University of Amsterdam, combining deep learning with music. Worked at OLX, Capgemini, and the Alan Turing Institute.

Outside of code: house and disco music as Crimson Lake, Berlin’s Floppy Disco collective, and photography.

About

Everyone says you need Python to build serious AI apps. That used to be true for training models, but most teams today are building products on top of APIs. This talk shows, with real code and production experience, why Ruby can be a better default for that work: less ceremony, cleaner agent abstractions, and faster iteration.

The thesis is simple: when complexity drops, product quality goes up.