The future of flavor discovery & development: Big food data and AI

The future of flavor discovery & development: Big food data and AI

By Subhra Pradhan

What is artificial intelligence?

Artificial intelligence (AI) can be simply described as the development of computer programs/systems that are able to perform tasks that usually require human intelligence. AI provides intelligent predictive solutions to problems based on existing data, inferences, patterns, and learning. The net worth of AI is predicted to reach $54 billion USD by 2026. AI is being applied to every field possible, from healthcare, agriculture, and education to aviation, finance, and market analysis. AI is also bringing in significant changes in the food industry, especially in product development.

Can AI accelerate flavor discovery and development?

What is flavor? Flavor is the distinct impression of food or drink that is perceived through our senses of smell and taste. The flavor compound (chemical profile) of an edible ingredient often largely determines how it can be used in combination with other culinary elements. While most of the AI platforms for flavor development that have recently emerged are using data from food and analytical chemistry, sensory sciences, and the aroma profiles of ingredients, some are using data on customer preferences, reviews, social media posts, and other inputs.

It is really exciting that AI and machine learning are being utilized to discover and develop new flavors and food combinations for human consumption. Given the fact that a newly discovered flavor with potential market value can take years to be developed into a new product, AI-based technology and analytics can provide the much-needed impetus to excel in competition. The food flavor market size is expected to grow to $12.8 billion USD by 2023. It is interesting to note that there are approximately ~106 recipes in use worldwide, compared to the vast theoretical possibility of > 1015 recipes. Hence a lacuna exists, and new combinations of food to be discovered through AI is one exciting avenue.

What’s in the market?


Foodpairing is primarily a Belgium-based flavor discovery platform that was co-founded in 2009 by Johan Lagenbick, Bernard Lahousse, and Peter Coucquyt. Their AI technology is based on principles of data science, physics, and chemistry. They have created one of the world’s largest databases of ingredients and flavors. Their long-term goal is to provide completely customized food solutions to promote healthy and sustainable living. Currently, their service has four plans, from free to expert ($49.99 USD per month), that can be used to develop unusual combinations of personalized flavors for food and cocktails. As it is thought that 80% of what we taste comes from aroma, the science behind Foodpairing starts with determining the odor profile of each culinary component through gas chromatography coupled with mass spectrometry. Next, the scientists match the ingredients through algorithms based on data science and machine learning. For instance, they found that strawberry is related to parmesan through cheesy flavor and to basil by citrus aroma.

Many innovative chefs and bartenders like taco man Wes Avila, Caribbean chef Jason Howard, and many others are creating unique foods and drinks using the Foodpairing platform. In addition to discovering new flavor combinations, the technology has been extended into the field of market research analysis to predict which flavor can create a successful business.

Plant Jammer:

Plant Jammer is a Copenhagen-based food tech startup founded by Michael Haase in August 2016. The science behind it is again food pairing based on aroma profiles, with the major difference being that this technology is for developing environmentally friendly plant-based food with the motivation of sustainability, resource efficiency, and reduced food waste. Most interestingly, its Gastro Wheel can formulate the recipe based on your choices and selections. Currently, the technology can be used as free web-based software, or it can be used as an Android/iOS mobile application. It is user friendly, allowing users to whip up a recipe in minutes from the ingredients available in their own kitchens. Before writing this article, I myself made a recipe using the Plant Jammer software with broccoli, asparagus, soy sauce, and almonds. I think it is a unique combination of flavors. Simultaneously, the software lets you avoid allergens and use ingredients available in real time while keeping the principles of their algorithm. This is good software for vegetarians and people with allergies, as they can choose their own ingredients.


ONE is a partnership between IBM and McCormick. In collaboration with IBM, McCormick — one of the leaders in the flavor industry — is expected to launch their AI-enabled flavor discovery and development software by mid-2019 to develop spice packets containing one-dish recipe mixes.

Their initial recipe mix will include Tuscan chicken, bourbon pork tenderloin, and New Orleans sausage. It is expected that McCormick’s workforce will be able to quickly navigate the data on millions of flavor profiles using the AI platform developed by IBM. McCormick’s new seasonings are expected to hit the market by late spring this year. According to the company, while developing a new food flavor, most of the effort and time is invested in ideation and development; hence AI can be envisioned as the future of flavor automation. The AI developed by IBM for McCormick is applied to pairing flavors, quickly screening through thousand of ingredients, identifying patterns in flavor combinations, discovering alternative ingredients, and finally, developing novel system-generated flavor formulas. Furthermore, the algorithm developed by IBM will not only access the flavor databases but will also integrate sales and marketing data so the flavors can be adjusted to geographic preferences.

More competitors:

There are many other AI-based platforms on the market, such as FlavorWiki, which is based on sensory profiles, consumer science, and trend analysis. dishq uses Spoonshot AI to predict consumer taste and food trends. Tastewise uses data from online recipes, home recipes, social media, and real-time restaurant orders to aid in market-driven food decisions. Whisk’s food map correlates properties of culinary ingredients such as nutrition, flavor, and shelf life with possible buying options. And Tastry is using analytical chemistry and flavor preferences.

On the other hand, Vinify and TasteMap are working with the wine industry to provide AI-based solutions for customized taste preferences. Analytical Flavor Systems is working with an AI-based tool called Gastrograph to create personalized food based on one’s own taste.

Riana Lynn, the founder of Journey Foods, is more focused on developing plant-derived snacks packed with dense nutrition to ensure appropriate digestion, boosting immunity and healthy brain function. The company has already developed products like Strawberry and Chia Seeds as well as Mango and Cayenne Spice bites.  

AI is the future of food & flavor discovery:

It is clear that AI based technologies for food & flavor discovery are becoming more prevalent, one of the next big questions on the horizon  is if robots can be our next generation of taste makers. Only time can tell us how the flavor discovery landscape will be shaped with the use of AI. But surely, we can expect some significant changes in how we perceive food in terms of unusual flavor combinations and tailored recipes. Considering the advancements in this field we can assume that along with food & flavor development the implementation of AI may contribute to  boosting sustainable and environmentally friendly solutions to various aspects of the world food problem.

If you have any questions or would like to know if we can help your business with its innovation challenges, please contact us here or email us at

Never miss an insight

Get insights delivered right to your inbox

More of Our Insights & Work

Never miss an insight

Get insights delivered right to your inbox

You have successfully subscribed to our newsletter.

Too many subscribe attempts for this email address.