Jordi Sahis wants to make AI go down well with more people. He has created the first Artificial Intelligence Gin. It is the first distillate co-created using AI tools.
The latest edition to a portfolio consisting of Jodhpur Gin, Krova Vodka, Relicario and Doble 9 rum, Spanish spirits company Beveland Distillers has launched its pioneering premium $42 Aigin.
Described as the latest innovation embodying the distiller’s motto of ‘spirit of excellence’, Aigin is the result of Girona-based Beveland’s goal to ‘achieve new levels of efficiency and competitiveness’.
The development of the new gin is based on algorithms capable of analyzing more than two million combinations of ingredients, distillation techniques, and maceration methods.
Says Marketing Director Sahis: “The AIGIN project is not simply about writing prompts and creating recipes. We’re exploring the limits of AI across all departments, leveraging more than 10 different AI-powered tools, The idea of using artificial intelligence to create a gin arose from a combination of factors. My background in Artificial Intelligence, with a master’s degree from Microsoft two years ago, opened my eyes to the possibilities of applying these tools in our daily work.
“In that context, we initiated a continuous improvement plan in the company, where we analyzed how new technologies could optimize our processes and increase our competitiveness. From the marketing department, we identified gin as an ideal product to experiment with AI, since the complexity of its botanicals and formulations has a significant impact on the final result and the consumer’s perception. The botanicals that were chosen were juniper, lavender, hibiscus,s and lychee.
Initially, the proposals generated by the AI were not very interesting, were too basic, and did not have anything to support the suggestions. The turning point came when we decided to create our own database to train the tool with our reality.
“We collected detailed information on the best-known gins on the market: the botanicals and ingredients they declared in their presentations and on the labels. We worked together with Innova Market Research to build this database. Then, we ‘trained’ our own AI with all this information. Through specific questions, the AI-generated a table of possible combinations of botanicals that had never been used together.
“We worked on these combinations in our laboratory to develop them. Many of these proposals were analyzed and classified by our team. We wanted to discover, with the help of AI, a gin with its own identity, one that was recognizable; which ruled out combinations that did not meet the brief.
“Interestingly, the floral profile of the gin emerged naturally. We were not looking for a specific profile from the beginning. We let the tasting panel, through blind tastings, indicate which formula was the most attractive and, therefore, what the public might demand. The formula chosen by our laboratory technicians, based on the AI analysis, exceeded our expectations.”
However, the use of AI did not stop at the creation of the formula. It was used for the packaging design, brand creation, and communication strategy. This project is a pilot test to explore all the efficiency improvements that artificial intelligence can bring to each department of our company without losing the human touch.
The gin represents the balance between tradition, technique, and technology. “But Artificial Intelligence alone cannot produce it,” continues Jordi Sahis.
“The passion and experience of the distillery team are the true driving forces.”
