Client work
Pour AI — a sommelier with no attitude
Client work, designed and built by Riv Lux Media.
The barrier to wine is not information. It is embarrassment.
People don't ask because they're afraid the answer will come with a lesson attached, or a vocabulary they're supposed to already have. So they buy the second-cheapest bottle and stop thinking about it.
Pour AI's whole positioning is the fix: your personal sommelier, no attitude required.
Two ways in, and the second one matters more
Ask a question. The obvious path.
Or photograph your meal. That is the one that changes who will use it. Describing a dish accurately enough to get good advice already requires food vocabulary — which is the same barrier again, one step earlier. A photo requires nothing. Point, shoot, get an answer.
Multimodal input isn't a spec-sheet feature here. It is the difference between a tool for people who already know how to ask and a tool for everyone else.
Tone is the product
"No attitude required" is a constraint that has to be enforced in the output, not just written on the landing page. An assistant that answers a simple question with a paragraph about terroir has failed, however accurate the paragraph is.
That means the design work sits in what the model is not allowed to do: not lecture, not assume prior knowledge, not hedge into uselessness, not perform expertise. Warmth and brevity are functional requirements in this category, and they are harder to hold than correctness.
Scope is what makes it trustworthy
An assistant that will answer anything will eventually answer something wrong, confidently, in public. In wine that is worse than useless — a confidently wrong recommendation costs someone money and their evening.
So the interesting engineering is the boundary: what the assistant answers, what it declines, and what it does when it isn't sure. We build AI features in that order — decide what "correct" means and how to measure it, constrain the scope to something it can be right about, ship the fallback for when it isn't, and only then make it feel good.
What this demonstrates
Multimodal input used to remove a real barrier rather than to demo a capability; tone as an enforced product requirement; and scoping an AI assistant tightly enough that it stays trustworthy in public.
Last updated