I need an end-to-end system that watches public car-sale sources, spots new listings the moment they appear, checks them against each user’s saved filters, and instantly sends a push notification when there’s a match. Scope • Data ingestion: scrape multiple websites while also consuming any official APIs that exist. The web scrapers must rotate user-agents/proxies, normalise the data, de-duplicate identical cars, and funnel everything into a single store (PostgreSQL or similar). • Matching engine: run lightweight, real-time rules that compare fresh listings to each user’s criteria (make, model, year, mileage, price, location, keywords). • Notifications: deliver push alerts through FCM for Android and APNs for iOS, including deep links that open the relevant listing inside the app. • Mobile app: a simple, clean interface on both iOS and Android where users can – sign up / sign in – create and edit saved searches – browse current matches with images, price, specs and the original source link – manage notification settings. A single cross-platform codebase (React Native, Flutter, or comparable) is preferred so we move fast, but I’m open to your recommendation. Deliverables 1. Source code for backend (scrapers, API, matching logic, notification service) with automated tests. 2. Containerised deployment scripts (Docker-compose or Helm) plus README instructions. 3. Mobile app source code and a short build guide. 4. Postman (or Swagger) collection covering every API endpoint. 5. One live demonstration that shows a new listing being ingested, matched, and pushed to a test device in real time. Acceptance criteria • New listings from at least three major car marketplaces appear in the app within 60 seconds of publication. • User filters return accurate results with no false positives in a 50-listing test set. • Push notifications arrive on both an Android and an iOS test handset. If this stack and workflow sound comfortable to you, let’s talk details and timeline.