Motif

About

Motif turns instinct into something searchable. You pick six images you genuinely love — interiors, clothes, art, objects, anything — and a multimodal model reads the shared patterns. It returns a named vibe, color DNA, preference insights, and a private gallery of your visual evidence. From that profile you can search for objects and get taste-shaped shopping results. Guests get one free analysis and one free search; verified accounts unlock five of each per rolling week. The product loop is tight: editorial homepage → upload → analyze → My vibe → shop, with private Blob storage, owner-checked image serving, and anonymous-to-authenticated profile claiming.

Why I built this

I built Motif because people often know what they like when they see it, but cannot name or search for it. I wanted a personal product that decodes visual taste into a reusable profile — vibe name, traits, colors, insights — then turns that into shopping results that actually match instinct. It was also a chance to ship an end-to-end Next.js product with editorial design, private media, AI analysis, and conversion-aware usage allowances.

Lessons learned

I learned how to keep AI analysis grounded in a structured taste profile, serve private evidence images through ownership checks instead of public Blob URLs, and design an editorial UI with Motion that still respects reduced motion. Building anonymous cookie ownership that can later claim into Auth.js accounts, plus weekly usage allowances and consent-gated analytics, sharpened how I think about trust, conversion, and launch readiness on a real consumer product.