Dating apps match people on photos and hope chemistry follows. But the dates people actually remember are taste dates — the right taqueria, the right film, the right record playing when dinner runs long. Suitor's concierge was built to stop guessing at that part. What it does Suitor is a live, token-based chat app (ladies chat free and invite-only; gentlemen spend tokens per message — burned on send, never paid out). Its AI concierge drafts messages for you — you always review and tap Send yourself. Taste Match is the concierge's date-planning brain: when you're ready to ask someone out, it extracts the interests from both bios, resolves them to Qloo entities, and asks Qloo's Insights API which places, movies, and artists people with that combined taste actually love. The result is a "Taste Match" card — a restaurant, a film, a soundtrack — each pick carrying Qloo's affinity score, plus a one-line date plan written around those picks. How Qloo is used • GET /search resolves bio interests ("vinyl", "tacos", "horror films") to Qloo entity IDs • GET /v2/insights with signal.interests.entities returns affinity-ranked place, movie, and artist recommendations • Cross-domain affinity is the point: the same taste signal that picks dinner also picks the movie and the music — no other recommendation source in the app can do that from two short bios Challenges & design honesty The concierge only drafts — it never sends a message and never moves tokens; tokens burn only when the user taps Send. Taste Match is read-only taste data. With no Qloo key configured, the feature degrades gracefully to a built-in taste bank, so the live demo always works.

Share this project:

Updates

Submission history