Closet

The Overview.

Closet is an ambitious fashion e-commerce platform built around a central premise: that AI can do what personal stylists do, at scale. The founding team came to Drille Digital with a product concept and technical roadmap but no design language, no brand, and no user experience framework. Our role was to take the idea from concept to a fully realized digital product, building the brand, designing the end-to-end user experience, and developing the web platform that would serve as the commercial foundation for the business.

The Challenge.

AI-powered fashion tools had existed in some form for years, but most had failed to cross the adoption threshold because they felt either too clinical or too gimmicky. The challenge was to build something that felt genuinely useful, where the AI layer enhanced the shopping experience without making users feel surveilled or reduced to a data profile. Trust was the central design problem: users needed to feel that Closet understood their style without feeling like Closet was watching them. On the brand side, the fashion market is brutally competitive and intensely aesthetic, and a visual language that felt generic or derivative would kill the product before it launched.

The Solution.

The brand we developed positioned Closet at the intersection of curation and intelligence, not a robot stylist, but a service that combined editorial sensibility with personal data. The visual language was clean and tactile, drawing from high-fashion editorial codes but keeping things approachable. The UX was built around a style profile onboarding flow that gathered preference data through choices that felt like editorial content rather than a form. The AI recommendation layer was surfaced through a conversational interface embedded naturally in the shopping journey, so it enhanced rather than interrupted the experience.

The Results.

Closet launched to an early access waitlist and exceeded its initial sign-up target within the first two weeks. The onboarding completion rate was notably high for a flow that collected as much data as it did, which validated the editorial framing approach. Early user feedback highlighted the recommendation quality and the visual experience as standout features.