Computational Flavor Science · Est. 2017
We build AI tools to explore how components of flavor combine to create delightful multi-sensory, chemical, emotional, and cultural experiences.
About
Since 2017, the Flavor Genome Project has been developing AI tools to explore how components of flavor combine to create delightful multi-sensory, chemical, emotional, and cultural experiences.
We process billions of data points across multiple large databases: large-scale recipe data, natural language and quantitative recipe reviews, ingredient data (including flavor compounds, nutrition information, and dozens of flavor and texture attributes), historical chef preference data, and ingredient adjacency data for global recipes.
The Flavor Genome Project was founded by Dr Altringer Eagle as a Lecturer at Harvard University's School of Engineering and Applied Sciences. Dr Eagle is now Professor of Design Engineering at Dartmouth College's Thayer School of Engineering, where she directs the Design Initiative. FGP is owned and operated by Ames Studio. For API access, consulting inquiries, or collaboration: info@flavorgenomeproject.com.
What We Do
We use natural language processing to identify data-driven patterns in how everyday people and experts describe flavor experiences, moving beyond conventional categories like cuisine type to discover new 'genres' of flavor experience.
We computationally understand how people experience flavor—what it tastes like, how it's described, how it feels—and match this to data on flavor compounds and nutritional information to generate novel, evidence-based pairings.
We automate the understanding of what people intuitively search for in a food or drink experience, regardless of the language they use, intelligently understanding flavor goals in context and making it easier to discover experiences they are likely to enjoy.
Forthcoming
The first notation system for flavor perception. Drawing on a decade of research from the Flavor Genome Project, Flavor Fluency translates perceptual learning into a teachable system—giving readers the vocabulary, frameworks, and creative practice to develop their own taste as a form of trained expertise.
Penguin · 2027
Origins & History
FGP's founder became a launch partner for Off Their Plate, a COVID-19 relief initiative that raised $6.5 million in four months to restore restaurant workers' wages by preparing meals for emergency workers. Press in the New York Times, Boston Globe, NBC, and Seattle Times.
Paired wines with new dishes by 25 top chefs for a Boston-based pop-up concierge delivery service, bringing chef-crafted dishes with FGP-generated beverage pairings to customers' doors.
The first public product powered by FGP data. Chef League was the first iOS game to let players compete on making things taste better—scoring millions of flavor combinations in real time. Players competed to fix recipes and master the functional use of salt, sweetness, acidity, fat, and spice. Press in VentureBeat, Harvard SEAS, Tapsmart, and Hackernoon.
Debuted a short film featuring five Boston chef-owners exploring how they 'design' flavor experiences. Sponsored by Harvard University at the MIT Cambridge Science Festival. Guests experimented with acids, umami, fats, sweetness, and bitterness through curated snack and beverage pairings.
Collaborated with the MIT Museum for the sold-out Food for Thought exhibition. Software-generated pairings like Porter and Stout with chocolate ganache and peppercorns surprised guests, deviating substantially from conventional pairing wisdom. Blind parallel tastings showcased how wine, beer, and cocktails interact with the same dish in different ways.
Blind pairing dinner with Chef Tracy Chang of Pagu. Sommeliers and software each suggested pairings for a seven-course menu. On control courses, guests could not distinguish between them. On courses with strong preferences, the software won both. Final scores: control (2 courses) tie; split (3 courses) tie; strong preference (2 courses) software wins.
Early software collaboration with chefs Tse Wei Lim and Diana Kudajarova, testing whether food and wine pairing can tell an emotional story. The collaboration demonstrated that the approach could inspire novel professional culinary designs.
Began building an interactive catalogue of 200,000 multi-sensory pairings across foods, beverages, emotions, and more. By 2019 the project would be processing over a billion data points.
Through a design residency in Italy, explored approaches to deconstructing flavor experiences with the Sensory Composition project.
The initial idea for FGP has roots in training for international tasting competitions with the Cambridge University Blind Wine Tasting Society. In 2008, Ames was a top-four finalist for the Lafite-Rothschild competition and won the 2009 Bollinger competition. The original inspiration: extending human memory, because professional tasting training helps people intuitively experiment with new flavors, but it is difficult to remember all possible flavor data.
Get in Touch
The Flavor Genome Project has grown into a large resource that enables various forms of flavor scoring and analysis for clients. Client services help fund and grow the project.
For API access, consulting inquiries, or collaboration, get in touch: