Sub-Society: AI Powered E-commerce store.
[Sub-Society] is an AI-powered fashion e-commerce app. Users post fits, earn levels, and tag buyable items; brands source creators directly. On LiDAR-equipped phones, a quick body scan captures key measurements. Smart wardrobe lets people build/share outfit like playlists while ML learns preferences to curate a store, and capsule suggestions. An in-app assistant recommends if outfit fits the wardrobe, plans event/vacation outfits and packs light using user constraints, itinerary, and owned items.
6 Months
2024
Digital Art
Team
Seymur Mammadov
Challenge
Core UX challenges: motivating content contribution, overcoming cold-start data sparsity, and merging social discovery with frictionless commerce without bias or overwhelm. We studied designers and everyday shoppers in Florence, ran interviews, mapped journeys, and audited social platforms and stores. We designed incentives (leveling, brand access), wardrobe capture flows, and explainable recommendations. Information architecture and feed/wardrobe/store triads were iterated to balance variety with focus, and to keep privacy, moderation, and ethics central.
Results
Deliverables included high-fidelity prototypes of feed, wardrobe, and store; an influencer marketplace; explainable recs; and an AI stylist for events and packing. Evaluations with fashion-forward and mainstream users reported easier discovery of indie brands, faster outfit decisions, and higher purchase intent (“it knows what I want”). Creators valued leveling and direct brand access. The wardrobe playlists improved reuse of owned items, reducing choice overload while keeping exploration fun.
Final Product
Product Development
Conclusion
[Sub-Society] suggests that coupling social creation, wardrobe intelligence, and transparent AI can make fashion shopping more personal, sustainable, and learnable. The project reframed personalization as a two-way dialogue: users express style via posts and outfits; the system explains and adapts.