TrueFind SG

Problem

Counterfeit and misleading listings are everywhere across Southeast Asian marketplaces, and shoppers have no consistent way to judge authenticity — every platform looks the same, and the signal that a listing is fake usually arrives too late. TrueFind SG unifies search across the major marketplaces and puts a trustworthy authenticity signal next to every result.

How it works

• Unified search across Shopee, Lazada, and Amazon.sg from a single query.
• Every listing gets an Authenticity Badge and Confidence Score, computed from AI analysis, official government scam advisories, and crowd-reported data — three independent signals rather than one.
• Image-based verification: upload a product photo for an instant authenticity read.
• A community flag system (upvote / downvote) feeds back into ranking to downrank suspect listings.
• A personalized dashboard to track verified purchases and manage flagged items.

Built cross-platform with React, Next.js, Expo, and Tailwind on a Firebase backend (Auth, Firestore, Cloud Functions, Hosting), with AI/ML models for scoring and image verification and Socket.IO for real-time updates.

Key decisions

Authenticity is a ranking signal, not a hard filter — suspect listings are downranked rather than hidden, keeping the catalog complete while steering shoppers away from fakes. Fusing three independent sources (AI, official advisories, and community reports) means no single signal can be gamed, and one Firebase backend serves both the web app and the Expo mobile client.

Recognition

Selected as a DSTA BrainHack 2025 Finalist.