AI that actually understands taste
Not just popularity rankings in disguise. Our recommendation engine combines collaborative filtering with content analysis to find hidden gems matched to you.
Two engines, one goal
Collaborative Filtering (ALS/SVD)
This approach finds users with similar rating patterns and recommends items they loved that you haven't tried yet. The math behind it — Alternating Least Squares and Singular Value Decomposition — creates a multi-dimensional map of taste. Users who consistently rate the same things similarly end up close together on this map.
The result: if someone with eerily similar taste to yours rated an obscure film a 9/10, that film jumps to the top of your recommendations — even if only 50 people have ever rated it.
Content-Based Analysis
This engine analyzes the attributes of items you love — genre, themes, mood, pacing, complexity, artistic style — and finds other items that share those qualities. It's especially powerful for discovering items in categories where you have fewer ratings.
Love slow-burn psychological thrillers? The content engine understands that preference and can recommend matching films, books, or games — even crossing category boundaries.
Why CatNLog recommendations are different
No popularity bias
We don't just recommend what's trending. An indie game with 100 ratings can rank higher than a AAA title with millions — if it matches your taste.
Cross-category intelligence
Your movie ratings inform your book recommendations. Love dystopian films? You might get dystopian novels too — because taste patterns transfer across media.
Gets smarter with every rating
Every rating you add refines your taste profile. The model continuously learns, so your recommendations evolve as your taste does.
Ready to discover your next obsession?
Join the waitlist and be the first to experience AI-powered entertainment recommendations.