TasteTwin

TasteTwin builds a compact “taste profile” from what you actually finish, skip, rewatch, and rate across movies, series, music, podcasts, books, and games. Instead of generic genre tags, it learns your preference vectors (pace, tone, complexity, humor style, violence tolerance, novelty vs comfort, etc.) and matches you to a small set of high-similarity “taste twins.” Recommendations come from (1) twin activity, (2) your personal constraints (time available, mood, device), and (3) a transparent explanation of why each item fits. The app avoids the usual endless feed by giving a daily short list (3–7 items) and a “one-click swap” if you’re not feeling it. It supports opt-in imports from services and manual logging for anything else. Realistically, the main challenge is data access and user motivation, so the product must feel useful within 5 minutes.