Skip to content

Date someone with actual taste

Your saved restaurants, your cuisine obsessions, and the hole-in-the-wall places you swear by tell us exactly who you are. Everyone says they "love food" - your dining data proves it.

Try the demo

How it works for food people

Bring your lists. Import your Google Maps saved places, or add your Beli, Tripadvisor and OpenTable favourites - we read your restaurant saves, ratings, cuisine preferences, and dining frequency. Your bookings stay private. Only taste signals are used.

The matching engine understands food at multiple levels:

Niche cuisines score higher

Everyone likes Italian. That is presence, not a personality trait. Cuisine rarity drives the niche weight.

Mainstream
Italian
Popularity 90 → NicheWeight 0.10
Niche
Georgian
Popularity 4 → NicheWeight 0.95 (capped)

Ratings reveal real preferences. Both saving the same restaurant is presence. Both rating it highly is genuine love. If you gave a place 5 stars and they gave it 2, the overlap disappears. The algorithm uses rating alignment as a sentiment factor.

What gets scored

Shared restaurants weighted by how niche they are - a shared love for a 30-seat ramen place scores far higher than both liking Nando's
Cuisine breadth and depth - not just "likes Asian food" but specifically Szechuan, Cantonese, Vietnamese, Korean BBQ - the granularity matters
Rating alignment on shared venues - the sentiment layer filters out places one person loved and the other thought was mediocre
Price bracket alignment - are you both street-food-first or fine-dining devotees? Lifestyle compatibility
Dining frequency - eating out 4 times a week vs once a month are very different lifestyles
Shared dislikes and dietary alignment - both avoiding chains, both vegetarian, or both allergic to the same things. Practical compatibility
Neighbourhood overlap - frequenting the same food scenes suggests you already move in similar circles

Your reservation for two

"You share 5 saved restaurants with a combined popularity under 8%, your cuisine profiles overlap 81% on Southeast Asian and Middle Eastern, and you both rate street food spots an average of 0.6 stars higher than fine dining. Your price bracket alignment is 94%."

Not "you both like food." The actual dining data.

The algorithm understands food obsessives

It knows that someone who has saved 40 restaurants across 12 cuisine types has a different relationship with food than someone who eats at the same 3 places. Engagement depth and cuisine diversity combine to paint a real picture.

The cross-category signal is strong too: if you also share music taste and attend food festivals together on Eventbrite, those signals compound.

Integrations

Beli
Manual entry · Restaurant saves, ratings, cuisine preferences, lists
Google Maps
File import · Saved/starred restaurants, reviews, visit frequency
Tripadvisor
Partner access · Restaurant reviews, ratings, cuisine tags
OpenTable
Partner access · Booking history, cuisine preferences, dining frequency

All opt-in. All revocable. Your bookings and locations are never shared - only the compatibility signals. Live links use an official API you authorise; file imports use the data export the service gives you; partner access needs the service's agreement; anything else you add by hand.

Affinity Atlas is in development

No matching is live yet. If you think restaurant compatibility should be a dating signal - or you have opinions about how cuisine rarity should weight matches - get in touch.

Stay in the loop