WebshopR

WebshopR is a social shopping app where discovery is driven by other people rather than by search. Shoppers browse products, vote on them in head-to-head “This or That” comparisons, share finds with friends through deep links, and use an AI shopping agent to find things by describing them rather than filtering. Products are synced from Shopify stores, so the catalogue is real merchandise from real merchants. It's built for people who shop the way they browse social feeds: opinion-led, sociable, and driven by what other people are picking.

Online shopping is built around knowing what you want. Search boxes, filter panels, category trees, all of which work well when you're replacing a kettle and badly when you're browsing. The way people actually discover things is social: someone shows you something, you ask a friend which one, you see what others picked. That happens on Instagram and in group chats, entirely disconnected from the stores selling the products, so the moment of interest and the moment of purchase live in different places. WebshopR pulls the social layer into the store itself. Voting turns browsing into a game, sharing works as a real link back into the product, and the AI agent lets someone describe what they're after instead of translating it into filters.

WebshopR is Nicolas D'Offay's concept. I was the sole developer, building it in FlutterFlow on Firestore with Algolia for search and Cloud Functions in Node.js for the backend. The pieces I'm most pleased with are the Shopify sync pipeline, where a Firestore-backed category mapping table replaced hardcoded logic so merchants can be onboarded without a code change, and the “This or That” voting and stories system, which needed custom FlutterFlow widgets and Dart actions to get the interaction feeling right. I also built the AI shopping agent as a three-webhook architecture across n8n, and the share and deep linking infrastructure that makes a shared product open to the right place in the app.