We built the tool we kept looking for
Atorie started with a frustration: every discovery platform kept surfacing the same labels, regardless of how specific or unusual the brief. We built a different kind of index.
A stylist, a product designer, and an ML engineer
Redouane spent more than eight years as a freelance wardrobe stylist in Los Angeles. A large part of the job was research: finding independent labels that matched a client's aesthetic, price point, and body type, when every shopping platform kept sending him back to the same twenty names. He kept a manual spreadsheet. It grew to 800 labels over the years.
In early 2024, he described the problem to Mia, who had been working on recommendation systems for a retail platform. She recognized the structural reason for the failure: purchase-frequency training data systematically underrepresents low-volume labels, regardless of quality. The algorithm is not broken. It is doing exactly what it was designed to do. That design is the problem.
Stefan joined to build the indexing and matching infrastructure. The three of us spent six months on the attribute taxonomy before writing a single line of the product. Atorie launched its index in January 2025 with 400 labels. It has grown since.
Three people, one very specific problem
Mia Tanaka
Background in consumer product design and personalization systems, including recommendation interfaces for style-adjacent consumer apps. She handles product design, the attribute taxonomy, and the matching model. She runs the Atorie style profile for herself twice a year to test whether it catches anything new.
Redouane Ramdani
More than eight years as a freelance wardrobe stylist in Los Angeles. He manages the label index, editorial review, and the designer relationships. He is also the reason Atorie exists: the tool started as a frustration he had been carrying since his first season.
Stefan Voicu
Builds the indexing pipeline, matching infrastructure, and the designer dashboard. His background spans recommendation and retrieval ML for consumer products, which gave him a useful frame for matching preference signals to low-volume catalogs at scale.
Independently made. No pressure to recommend what pays.
Atorie is bootstrapped. There are no investors telling us to surface labels that advertise. If a label appears in your feed, it earned the placement through style-profile match, not paid placement. Start your profile, or reach out if you want to list your label.