Tinder for Restaurants
Tinder for Restaurants

Tinder for Restaurants

NOMNOMNOM is like tinder for restaurants

  • Finding what you want it difficult: you are looking for a restaurant but which one?
  • There are so many options, around 4000 in Amsterdam. Something hip or someplace cozy?
  • Plenty of platforms make you rely some influencer's (bought and paid for) taste, but how do you explore your taste?
  • NOMNOMNOM assumes you don't already know exactly what you are looking for but you'll recognise what you like and don't like.
  • This AI offers you some suggestions after which you rate them, at each step it shows you more of the things you like and less of the things you don't, while surprising you along the way.

How does it work

  • NOMNOMNOM is powered by a discovery engine, a hybrid search and recommendation algorithm, and is trained by learning that similar users have similar taste and like similar restaurants.
  • The engine is build in python and extends the most cutting edge recommendation systems to allow for interactive, session based product discovery.
  • The frontend is build in React, allowing for a dynamic and responsive user experience on both mobile and desktop.

Stepwise discovery use cases

The NOMNOMNOM discovery approach generalises to e-commerce where taste matters. The same system, swipe products left or right to like or dislike, find what you want in in steps while balancing the familiar with the unexpected.

  • Fashion, socks and shoes
  • Toys & Gifts
  • Accessories, like bracelets and jewels
  • Food, restaurants, delivery & takeout meals