StartAppLab

Pickwell is a barcode-first, AI-powered product intelligence platform that makes food labels easier to understand and transforms individual product scans into shared, reusable knowledge. The idea was inspired by a problem close to home: the growing burden of chronic kidney disease (CKD) in the Philippines, where understanding nutritional information, particularly sodium content, can matter greatly. Yet everyday consumers often face tiny nutrition labels, unfamiliar ingredients, and confusing packaging. Pickwell makes product information more accessible. Users simply scan a product's barcode. If the product already exists in our shared database, its information appears instantly without requiring another AI inference. If it's unknown, users photograph the front packaging, nutrition facts, and ingredients. Powered by Meta Muse Spark 1.3, Pickwell interprets these images together, extracts structured product information, explains ingredients, highlights nutritional considerations, and communicates uncertainty. Users can immediately explore their own private analysis. But Pickwell goes beyond image recognition. Our Product Trust system checks the extracted evidence for inconsistencies and prevents unreviewed submissions from automatically entering the shared catalog. Human-reviewed, approved products can then become available to everyone who scans the same barcode. Built with Flutter, Supabase, and Meta AI, Pickwell combines multimodal reasoning, deterministic validation, evidence-bound publication, and shared caching into one consumer-focused platform. Our vision is simple: one scan helps one person understand a product; one trusted contribution can help countless others. Pickwell isn't just about reading labels. It's about building a collaborative foundation for more transparent, accessible product information, starting in the Philippines and designed to grow beyond it.