Phantasm & Brain Applied AI Product Lab
Coming soon · tasteinmotion.app

It understands your taste, not your history.

Taste in Motion is an intelligent B2C recommendation engine that understands nuance. By analysing user preferences across disparate categories, from fine wine to niche perfumery, our algorithm maps personal aesthetic profiles to deliver highly accurate, cross-category recommendations.

At a glance
  • Category
    B2C · Cross-category recommendation
  • For
    People who know what they like but not what to buy
  • Signal
    Aesthetic profile, mapped across unrelated categories
  • Status
    In development. Early access list open
01The problem

Recommendation engines don't know you.

They know what you clicked, and they know what people who clicked the same thing clicked next. That is a description of a crowd, not of a person.

So the recommendations arrive category by category, with no memory across them. The system that knows your films has never met the version of you that buys perfume. The one that knows your wine has no idea what you read. Every category starts you over from nothing, and each one flatters the average rather than the individual.

But taste is not category-shaped. The person who loves a particular kind of film very often loves a particular kind of whisky, and the reason is legible, provided you are modelling the aesthetic underneath rather than the purchase on top.

02 How it works

Map the profile. Cross the categories.

  1. Read the preference, not the purchase

    The engine starts from what you actually respond to across the categories you already care about, and from the things you reject, which are frequently more informative than the things you like.

  2. Build an aesthetic profile

    Those responses resolve into a profile that describes the shape of your taste rather than a list of products: what kind of thing moves you, and why.

  3. Translate across domains

    Because the profile is category-independent, it can be applied to a category you have never touched. This is where a cross-category recommendation stops being a guess and becomes a translation.

  4. Explain the reasoning

    Every recommendation carries the connection it was drawn from. A suggestion you can interrogate is a suggestion you can trust, and correct.

03What makes it different
01

Cross-category by design

Most engines are built inside one vertical and stay there. Taste in Motion treats the movement between categories as the product, not an edge case.

02

Nuance over popularity

Ranking by what sells most returns the same answer to everyone. The profile is built to find the specific thing, including the obscure one.

03

Legible recommendations

Every pick comes with the reasoning attached, so the relationship is a conversation you can steer rather than a feed you scroll.

Early access opens soon.

Taste in Motion is in active development. Tell us which categories you would want it to understand first and we will bring you in early.