UST Uses AI to Help With Beverage Choices

You’ve turned over so much of your daily life to computer gadgets—from monitoring the doorbell to starting the car. Now, well, how about letting machines pick your wines and beers?

UST in Aliso Viejo, one of OC’s largest privately held tech services firms, is helping create your virtual sommelier. It has launched UST AiSense, which provides personalized food and beverage recommendations for wine, beer and some foods, all using what is called “sensory science.”

The digital, artificial intelligence technology is a “palate-based recommender” that helps consumers select the products they will enjoy.

The program starts after you scan a QR code, and take a brief palate-based, individualized survey: 20 questions to identify your personal taste and flavor—such as how do you feel about the taste of sweetened coffee.

A “perfect palate” will be matched in 20 seconds, according to the company. The technology provides recommendations based on your inputs, and matches specific wines based on those preferences. It also recommends “associated foods” to be paired with the wine.  

Users then get wine or beer recommendations and suggestions for foods to accompany them.

Sensory Sciences

UST AiSense combines an analytical chemistry framework together with artificial intelligence/machine learning to analyze the chemistry of the sensory-based products to evaluate the quality and predict market performance of products, such as beverages, and food, said Mahesh Athalye, senior director of the company’s retail platform and solutions.

It utilizes technology developed by Tastry, said to be the world’s first AI-driven sensory sciences company. Its technology combines analytical chemistry, consumer flavor preferences, and machine learning to predict market performance for sensory-based products.

UST is a strategic investor in Tastry.

Athalye said the data is also provided back to producers and manufacturers to enable them to integrate sensory information during product development. For example, for wine, the vendor’s computational blending process can help to predict the likeability of various blends for retail selection.

“The software will recommend what food goes well with specific wine that is recommended. Again, all based on the individual inputs gathered in the initial survey,” Athalye told the Business Journal earlier this month.

For example, the system recommended for him a dry, medium bodied crisp chardonnay— a 94% match based on his initial inputs of taste preference.  

Along with that, the consumer receives suggested pairings with the wine—hard cheese, smoked poultry, fried shellfish, etc.

“With this information, retailers can truly optimize their product mix based upon consumer preferences, stocking shelves with the exact products customers want,” Athalye said.

Sales Boost 

UST says the technology can increase category sales by 10% to 15%, increasing the value for retailers.

Retailers can offer UST AiSense recommendation solutions within a store kiosk or through a website or a mobile app. The system provides real-time insights on sales, inventory, and customer preferences.

“Each person has their own unique taste preferences. You no longer need to spend time researching the right wine and end up buying something that doesn’t match your specific taste,” said Keith Pickens, retail domain leader–general manager at UST.  

The mobile app can be customized for specific client’s store location, be it retailers, convenience stores, or restaurants, the company said.

The system also aims to let consumers be adventurous by guiding them to try beverages and foods the AI knows one will probably like but may not have tried before.

“With UST AiSense Solution, the food pairing capabilities scale from basic food categories (meats, cheeses, etc.) to specific prepared foods and recipes, identifying the very best wines that pair both with food choice and shopper’s palate. Retailers can use this feature for cross selling and upselling products,” the company said.