Money-Saving Shopping Companion App

Customer: AI | Published: 23.02.2026

I’m building a cross-platform companion that cuts everyday shopping costs in just one tap. It has to run natively on both iOS and Android and feel identical on each. Core idea • After every shop I’ll photograph or scan my receipt; the app needs solid OCR and SKU matching so it can learn what I regularly buy and flag cheaper alternatives next time. • Before I head out, I want to create a list by typing, speaking, or simply snapping pictures of products. Using geolocation, current store catalogues and real-time social-media promos, the app should map the best nearby store within the distance range I set. • Coverage isn’t limited to groceries. It must work just as well for clothing and household items, recognising fashion pieces from user photos and pairing them with the best available offer. • Social-media scraping is essential: the moment a supermarket or fashion retailer posts a flash sale, the engine needs to pick it up and factor it into recommendations. Key deliverables 1. Universal iOS + Android build (Swift/Kotlin or Flutter/React Native acceptable if performance and camera features stay intact). 2. Receipt OCR module with price comparison engine. 3. Multi-mode list creator: text, voice (on-device or cloud ASR), and image recognition. 4. Offer aggregator pulling data from store APIs, public catalogues and social channels in real time. 5. Recommendation algorithm that ranks stores by savings, distance and user preferences, returning a clear shopping route. 6. Admin dashboard or endpoint so I can tweak store feeds and distance thresholds without redeploying the app. Acceptance criteria • Receipt scan must extract at least 95 % of line items correctly on a standard supermarket docket. • For a 10-item mixed list the app returns results in under 5 seconds and shows minimum 3 cheaper alternatives when they exist. • Push alert fires within 2 minutes of a monitored retailer posting a relevant deal on social media. • All three input modes create an identical, unified list for processing. Code, brief setup docs and a short demo video will complete the hand-off.