Nuvo

like the French word for new, “nouveau”

Discover restaurants you’ll love, tailored to your tastes.

Problem: Paradox of Choice
The challenge for many food lovers is the daunting task of sifting through endless options to find restaurants that not only cater to their tastes and dietary needs but also fit their budget and ambiance desires. Traditional search methods and apps tend to offer a deluge of options, lacking personalization, which can lead to decision fatigue and dissatisfaction.

Solution: Instantaneous Personalized Discovery
Nuvo is a personalized, AI-driven restaurant discovery experience, tailored to match individual tastes, dietary needs, and location preferences with simple, intuitive onboarding.

Technical Details:

First, we input a user's preferences in the onboarding process.
For every restaurant that we have scraped into our database, we:
- create a collage of all of the images that we can find for that restaurant and save it as an image embedding
- scrape as much data as we can off of Yelp and other internet sources. We embed all of this text in the database.
When a user inputs a request, we use a vector search on the image embedding and on the embedded text data to serve the user a personalized recommendation.
We also integrate langchain to create an interactive chatbot-like interface for the user.
Stack:
- MongoDB, Python Backend with Langchain integration, NextJS Frontend

Major User-facing Components:

Onboarding for New User Experience
Search (prompt entry) + Location indicator + Feed of personalized recommendations
Profile, including preferences (to adjust the outcome of the NUX onboarding)

Future: multi-player discovery so a group of people can collaboratively search to find restaurants that match combined tastes.