Features · August 9, 2026

How Shelfies AI Recommendations Actually Work

Shelfies reads your shelf — ratings, genres, pace, what you abandoned — and recommends from that, not from what is trending. Here's how it works.

Most book recommendations are bestseller lists with your name on top. They tell you what is popular, which you already know, and ignore the one thing that actually predicts whether you will like a book: what you have liked before.

It reads your shelf, not the charts

Recommendations are generated from your actual library — what you rated highly, which genres you return to, how fast you get through them, and what you marked DNF. That last signal is the most underused in the whole category: knowing what you abandoned at page 60 says more than any five-star rating.

Match my mood

Some nights you do not want a good book, you want a specific kind of book. Pick a mood — relaxed, intellectual, adventurous, emotional, light and fun, or intense — and the suggestions narrow to that shape.

Ask it directly

You can also just ask: what is my longest book, which genres do I read most, what should I read after Dune, suggest my next read. It answers against your library rather than the internet at large.

Why this is better than a bestseller list

Because you are not the average reader. The value of a recommendation engine is entirely in how well it knows the exception — and the only way to know that is to have your ratings, your finishes and your abandonments in one place.

Shelfies is a modern, mobile-first reading tracker with AI recommendations, progress tracking, streaks, and quotes — built to replace Goodreads and StoryGraph for readers who want a faster, prettier experience.