Case study — E-commerce · Web
daily.style
An in-house fashion aggregator that gathered Bulgaria's online shops into one catalog.

Shopping fashion online in Bulgaria meant a dozen tabs across a dozen stores. We wanted one place where the market's catalogs met — and we wanted it to run itself. We built it in-house, from architecture to launch.
A Next.js platform on Vercel ingested product feeds through the Tradedoubler affiliate network, classified every item with AI embeddings, and admitted only true apparel through a 262-category ingest gate. The catalog refreshed itself on a daily sync — no manual curation, no stale listings.
A self-maintaining product: feeds arrived, the gate filtered, the catalog renewed daily — proof of the platform engineering we bring to client work. The engine — feed ingestion, product classification and visual search — is being carried into our e-commerce work.
Daily
Catalog sync
262
Categories in the ingest filter
500,000+
Products

Have a project in mind?
Selling fashion online? Get a free AI shopping readiness audit of your catalogue.
Request the free audit