🤖dVIN + AI

The future of wine education and curation

dVIN plans to use generative AI in several key aspects of the user experience on our platform, and we are exploring several ways to implement $VIN as an incentive to grow our language model.

Wine is a complex and often intimidating subject. Our aim is to use AI to provide easy and approachable information and education to users, that over time becomes personalized to their taste, style and sphere of knowledge.

Rather than display pre-written static wine information, we intend to use curiosity on the platform as a prompt to start a conversation with the user. For example, if a user taps on any of the following text in the app, we will open a chat window to provide a personalized, generative response, including prompting the user to continue the conversation:

Glossary/information

• Wine regions (i.e. Bordeaux, Napa Valley, Mornington)

• Appellations (i.e. Pomerol, Barolo, Eden Valley)

• Grape varietals (i.e. Merlot, Cabernet Sauvignon, Grüner Veltliner)

• Winery history (i.e. Pétrus, Sine Qua Non, Penfolds)

• Winemaker biographies (i.e. Bill Harlan, Peter Gago, Michel Rolland)

Subjective content

• Tasting notes (how should you expect a wine to taste)

• Vintage ratings (is the 2012 Pomerol ready to drink?)

• Food pairing suggestions (what to eat/serve with this wine)

Ratings and price comparisons (what do the professionals think of this wine, and how much should you pay for it)

When the user taps on any of these tools, the generative answer will be provided in chat, with a prompt to ask additional questions, dig deeper into the particular subject, or to make adjustments to the style and depth of the generative responses in the future.

The conversational nature of the AI will also serve as a non-intimidating, non-judgemental, always available educational source for budding wine enthusiasts who might not want to admit a lack of knowledge to a sommelier or wine shop attendant. Additionally, the AI has the opportunity to learn the users preferences and knowledge base each time it is used.

Additionally, the AI will proactively reach out to the user for:

• Feedback on food pairing suggestions after the fact

• User tasting notes for wines after a Digital Cork is opened or a Tasting Token is minted

• QA feedback after wine has been shipped and received

• Follow up after wine events and experiences

User responses will create a wine profile for each user that over time will be used to customize the user experience on the app, personalize the voice of the AI wine guide, propose wines to buy, and even complete purchases without ever leaving the chat.

Through our network, dVIN has access to a significant amount of proprietary wine content, from tasting notes and wine data to written wine journalism and scores all to serve as a base for the AI model.