Ferreter
Your taste, uncovered. A recommendation engine for film and television built on your soul, not the algorithm — a Claude-driven agentic loop reasons over your stated taste and prior signals to make picks a person would make.
Ferreter is a recommendation engine for film and television built around a simple premise: algorithmic feeds optimize for engagement, not taste. Ferreter optimizes for the latter.
Under the hood, a Claude-driven reasoning loop combines a user's declared preferences, watch history, and rating signal with structured metadata from TMDb to produce recommendations that feel deliberate rather than generic. The agent justifies each pick in plain language so users can trust — and steer — the result.
Built end-to-end as part of the Auxelion product line.