
PantryPal solves a critical problem: one-third of food bought for the home is thrown away, not from lack of motivation, but because people forget what they own. By the time they remember, it's expired and in the trash. Our AI solution inverts the traditional recipe app workflow. Instead of asking "what should I cook?" then shopping for missing ingredients, users first scan their groceries. PantryPal reads product labels via OCR to capture real nutrition data, allergen information, and expiration dates. The app then organizes their pantry by urgency, surfacing items about to expire first. When users ask "What can I cook tonight?", Claude AI generates recipes using only ingredients they already have—respecting hard allergen constraints and providing accurate nutrition data from actual packaging, not generic databases. Cooking a recipe logs it against daily macro targets, creating a closed-loop system that connects pantry to plate to nutrition tracking. Key features: Intelligent barcode/label scanning (3 photos for packaged goods, 1 for produce), allergen-aware recipe generation, real nutrition data per 100g, pantry organization by expiry urgency, integrated nutrition tracking. Built with React, Tailwind CSS, Supabase, Claude API, and AWS Bedrock. Target market: Home cooks managing dietary restrictions, fitness goals, or health conditions. Addressable market: 15M+ US households.
10 Aug 2026