Stockholm Company Launches A.I. Wine Guide in 19 Languages

A Stockholm-based wine technology company said it has launched an artificial intelligence service designed to help consumers identify wines, choose pairings and manage personal collections from a mobile phone.

The service, called VinSavant, was announced in a press release dated Oct. 4 and is available through the company’s website in 19 languages. The product is aimed at both new wine drinkers and more experienced users, including restaurant diners who want help reading a wine list and collectors who track bottles at home.

According to the company, users can photograph a wine label and receive an AI-generated analysis that includes a score, food pairing suggestions and recipe ideas. They can also search for a wine and get information about grape variety, region, ratings, aging potential and serving advice.

VinSavant said its system can also be used in restaurants. A user can scan a wine list and receive information about the wines shown, then photograph a menu to get pairing suggestions for each dish. The company said the process can also work in reverse, with users entering a dish name or scanning a menu first and then receiving wine recommendations.

The company was founded by Urban Petersson, who said he built the tool after more than 15 years of attending wine courses and tastings. In the release, Petersson said the service began as a personal project and was later shared more widely after experienced sommeliers started using it.

The company presents VinSavant as an alternative to wine services built mainly around fixed databases of reviewed or cataloged bottles. In the release, VinSavant said those databases remain useful but are limited to information that has already been entered by editors or users. It said its own service relies on AI to interpret a label, assess a wine’s likely character and retrieve updated information about vintages and producers.

That distinction reflects a broader shift in consumer software, where AI tools are being marketed as more flexible than traditional searchable catalogs. In VinSavant’s case, the company said the goal is not to replace wine communities or existing databases but to move from general opinions about a bottle to more specific advice, such as what food would pair best with a particular wine on a particular night.

The release also described a cellar management feature that allows users to organize bottles by storage location, drinking window, personal rating and notes. People who already keep collections in CellarTracker or Vivino can import them, the company said, after which the wines are automatically enriched with AI-generated data. Collections can also be shared with other users, who can be invited to view or edit them.

Other functions listed by the company include a recipe bank for saving and sharing dishes linked to searches and label scans, vintage comparison tools, and an AI chat function for general wine questions. The service is available at vinsavant.ai.

Petersson framed the product as a response to the complexity of wine and to the amount of information consumers face when trying to learn about it. In the release, he said wine books, databases, podcasts and television programs often produce information overload because the material is scattered and not always current. He cited the example of Italy, which he said has about 550 officially recognized grape varieties used for wine production, each affected by place, growing conditions, vintage and winemaking method.

He said artificial intelligence changed the way he approached those questions by allowing him to begin with a prompt and let the system locate updated information about a vintage or winery. In the release, he said that meant he no longer had to gather all the information in advance to have answers ready.

VinSavant’s announcement comes as AI tools continue to spread into niche consumer categories beyond general search and writing applications. Wine is a field where many products already exist, from label-scanning apps to collector platforms and professional review databases. VinSavant is entering that market by emphasizing multimodal use, including photographing labels and menus, and by promoting a more conversational form of guidance.

The company said the service is intended to help users answer practical questions such as what wine to serve with dinner, how long a bottle can age, or what to order from a wine list with many options. It also said the app can help users plan meals around bottles that are approaching their ideal drinking period.

The release did not specify pricing, subscription structure, mobile platforms or launch promotions. A placeholder in the announcement indicated that those details were expected to be added, but they were not included in the version distributed.

No user figures, revenue data or outside investment information were disclosed. The company also did not provide independent validation of the accuracy of its wine analyses or pairing recommendations. As with many AI products, the quality of responses is likely to depend on image recognition, source material and the model’s ability to interpret incomplete or ambiguous information.

Still, the announcement shows how developers are trying to bring AI into highly specialized consumer decisions that have traditionally relied on experts, reference books and enthusiast communities. In this case, VinSavant is positioning itself as a digital sommelier that users can carry in a phone, whether they are standing in a store aisle, looking at a restaurant menu or deciding when to open a bottle already in their cellar.

The company said the platform is already live and can be accessed through its website. Petersson, in the release, said he continues to attend wine courses and tastings even after building the product, describing the service as a companion to learning rather than a substitute for it.