Meet Dolores

Chicory Releases New Food AI, Machine Learning Engine

Advanced Technology Uses Consumer Behavior to Match Digital Recipe Ingredients to Products

Chicory, a New York-based foodtech company paving the way for grocery ecommerce, today announced the release of their artificial intelligence engine, “Dolores.” Dolores introduces advanced technology, which intelligently matches digital recipe ingredients to purchasable products that consumers are or will be looking for. Consumers can then buy the shopping list with just the click of a mouse via one of Chicory’s grocery integrations.  

“We were able to teach Dolores to implement the same subjective reasoning about food that humans make,” said Yuni Sameshima, Chicory CEO and Co-Founder. “For example, Dolores sees that the recipe requires pepper and can deduce whether that is black pepper or a bell pepper based on the context, which allows us to better provide users with their desired ingredients.”

The scale and diversity of the data that is incorporated in to Dolores makes the AI system completely unique and proprietary, and reflects actual user habits of the people that interact most prominently with recipes. With over three million recipes in the Chicory system, Dolores runs on information parsed from democratized recipe content from the Chicory network which was used to build a categorization system and product taxonomy. The final product identifies food brands, bundles related items and suggests related items to the consumer when an ingredient is out-of-stock. Dolores also matches recipe ingredients to appropriate products, and vice versa. Dolores can also determine whether the recipe it is reading calls for a specific product or brand or whether the user can go with an alternative, should the item not be available at their selected retailer.

Consumer behavior and interaction with Dolores will further inform its machine learning algorithm making it even smarter the longer it is in use. User-generated signals like product swaps in their carts will aid in supervised learning and show how grocery shopper habits are changing and ingredient preferences are developing.