ITS 4.0 · 2024 · ITS Agroalimentare del Piemonte
Foodpairing
A food-pairing database built from commercial data and 1,000 interviews, a RAG AI suggesting innovative pairings, and a free web app to explore with custom filters.
The story
The project, step by step
The project story
Foodpairing was born to turn the science of food pairing into a practical tool. The project aims to build a food-pairing database combining two complementary sources: structured data from commercial databases such as FlavorDB and human insights gathered through 1,000 interviews.
The AI that suggests
A RAG (Retrieval-Augmented Generation) AI is trained to analyse this data and improve the ability to suggest innovative, personalised pairings. Combining structured data with human insights produces a recommendation system that is more accurate and aligned with users' preferences.
The free web app
A free web app lets users explore flavour combinations through an intuitive interface and custom filters. The AI behind the app analyses users' choices, suggests surprising pairings and keeps improving thanks to the feedback it receives.
The results
The database keeps growing thanks to users, who contribute recipes and suggestions. This active involvement makes the culinary experience dynamic and evolving, supporting the mission of improving the food supply chain in terms of sustainability and innovation.
Photos
Images from the project
Slides
The project presentation
Video
Watch the project
Have a project in mind?
Tell us what you need: together we will design the right solution.