Our process
Our focus is futuristic silhouettes that feel both flattering and comfortable.
Every piece starts with an old-school sketch. Pen and paper. Sometimes watercolor.
The sketch is then transformed into editorial imagery with AI to imagine what the garment could look like IRL. We’ve developed a precise process for preserving the originality of the sketch while still making the result feel realistic. That’s surprisingly difficult with today’s image models, which often drift away from the original input in pursuit of “realism.”
From there, we iterate with AI until something clicks. Sometimes we intentionally let the AI hallucinate to surface unexpected directions. Whether those ideas make it into the final piece is ultimately a matter of taste.
Once we’re happy with a concept, we gather public input before selecting garments to produce. Each piece is made-to-measure, often through multiple fittings. Much of the construction is draped directly on a mannequin, though we also use AI-assisted pattern-making for more architectural silhouettes.
Software inspiration
Every yanabana fabric starts from the same base pattern inspired by mainframe computer punch cards.
The original computer punch cards actually came from fashion.
Early computers borrowed the idea from the Jacquard loom, a programmable weaving machine from the 1800s that used punched cards to control textile patterns automatically. Ada Lovelace even described computers as “weaving algebraic patterns.” So it felt right that an AI-native fashion brand should begin there.
Not with florals.
Not with monograms.
But with the origin of programmability itself.
We also love the aesthetic of the pattern itself: rigid grids, tiny encoded decisions, human creativity compressed into machine-readable form.
Most people won’t notice it at first glance. But hidden inside every garment is a reference to the moment humanity first started turning creativity into code.
