Água do Vimeiro: turning mineral water into a generative consumer experience.

The challenge: from a product to an experience

What happens when a consumer brand stops using artificial intelligence merely to communicate a product and starts using it to create new ways of experiencing it? That was the opportunity behind Mocktails Vimeiro.

Água do Vimeiro wanted to explore new consumption occasions around its mineral water portfolio while responding to growing interest in sophisticated alcohol-free alternatives. A conventional approach could have produced a catalogue of recipes. Instead, we built a system that could create them.

Developed by Algorithm G through GAIA^©, with nutritional curation by clinical nutritionist Mariana Chaves, Mocktails Vimeiro lets anyone describe the drink they want and receive a personalised recipe using Água do Vimeiro. A summer refreshment, a children's party drink, a lower-calorie option, ingredients already at home, or a dinner mocktail can all become different starting points.

From three flavours to thousands of possibilities

Consumer products have traditionally been built around finite choice. Generative AI introduces a different possibility: instead of asking people to choose only from experiences designed in advance, a brand can generate an experience around an individual's context.

As the campaign puts it: Life asks for much more than three flavours. Mocktails Vimeiro is not a recipe database with a conversational layer. The recipe itself is generated. Two people can use the same brand experience and leave with completely different drinks.

The system can incorporate Vimeiro Gás, Vimeiro Sparkle, or Vimeiro Original according to the character of the requested drink. The products become ingredients inside a dynamic service, not branding placed around an AI application. The website becomes somewhere consumers can do something with Água do Vimeiro.

Useful generation needs context, expertise, and constraints

Generative AI is exceptionally good at producing possibilities. Brands cannot operate on possibility alone. They have identities, products, communication principles, knowledge, and boundaries; when food, nutrition, or wellbeing are involved, those boundaries matter even more.

The challenge was not simply to generate a mocktail. A general-purpose model can already do that. It was to build a system capable of generating Mocktails Vimeiro consistently. Água do Vimeiro supplied the brand and product context, Mariana Chaves established the nutritional framework and ingredient universe, and Algorithm G used GAIA^© to translate those requirements into the architecture and behaviour of the application.

The experience follows four simple principles: 0% alcohol, 0% added sugar, 0% soft drinks, and 0% artificial sweeteners. Consumers can also ask the system to consider calories. The expert defines the territory; the AI explores it. Human expertise does not disappear. It becomes part of the infrastructure.

Building the generative layer with GAIA^©

GAIA^© is Algorithm G's proprietary system for designing, validating, and deploying production-ready generative AI applications. It creates the architecture between what a user asks and what the underlying AI ultimately produces.

For Água do Vimeiro, that architecture had to interpret open-ended consumer requests while remaining inside a defined brand, product, and nutritional environment. The consumer sees a deliberately simple conversational interface. Behind it sits a structured, multi-layer generative system combining instructions, contextual knowledge, model selection, action layers, iterative testing, and human validation.

GAIA^© has been applied across more than 80 live applications and multiple business contexts. With Mocktails Vimeiro, that complexity needed to remain invisible. The experience succeeds when people do not have to think about the technology. They ask for a drink.

“Building an AI that can suggest a mocktail is easy. Building an AI experience that consistently understands Água do Vimeiro, incorporates Mariana's expertise, responds to thousands of different consumer intentions and remains simple enough for anyone to use is the actual product challenge. That is exactly the type of problem GAIA was designed to solve.”

Gonçalo Perdigão, Co-founder and Managing Director of Algorithm G

From briefing to market in one month

The entire deployment moved from initial briefing to public launch in one month. That month did not cover technology alone. It included concept and experience design, GAIA^© generative architecture, nutritional knowledge integration, development, testing, optimisation, a public consumer experience, a multi-channel campaign, and promotional video content.

An internal proof of concept is not the same as a consumer-facing product ready for public launch. Technology had to coexist with brand, communication, content, expert knowledge, user experience, and a fixed operating deadline. AI became part of an integrated product and communication launch, rather than a separate long-term technology programme.

• Concept and experience design
• GAIA^© generative architecture
• Nutritional knowledge integration
• Development, testing, and optimisation
• Public experience and multi-channel launch

Generative personalisation, not segmentation

During the first month, consumers requested thousands of personalised mocktails. Each request represents something different from a conventional digital impression. A page view is content consumption; a generative request is an expression of intent. The person describes what they want and asks the brand to respond.

Traditional personalisation often means placing people into segments and selecting a predefined experience for each group. Generative personalisation creates the final experience when it is requested. The brand does not need to anticipate every flavour, ingredient, mood, or occasion in advance; the system dynamically creates the last mile.

That is where generative AI becomes strategically interesting for consumer brands. Using AI to produce an advertisement changes how a company communicates. Using AI to create a service changes what a customer can do with the company.

The product remains physical. The experience becomes software.

A bottle of Água do Vimeiro remains a bottle of mineral water. What changes is the layer surrounding it. Software can create dynamic context around physical products: a food brand can create a personalised cooking agent, a fashion company an interactive styling experience, and a water brand thousands of personalised drinks.

Mocktails Vimeiro can continue to incorporate different inspirations, seasonal contexts, and consumption moments. It becomes a persistent digital asset, not only a campaign. For Água do Vimeiro, it opens a new way to discover its portfolio. For Mariana Chaves, it shows how professional expertise can scale without removing the human expert. For Algorithm G, it demonstrates what GAIA^© was designed to do: turn a specific opportunity into a production-ready application that real people can use.

From briefing to market in one month. From one physical product portfolio to thousands of personalised interactions. A bottle of water cannot contain infinite drinks. The experience around it now can.