Iscanow is the project I'm most proud of, because it solves a problem I could actually see: small printing and stationery shops with hundreds of unorganized products and no fast way to price unfamiliar items.
The idea was simple — what if the cashier could scan an unknown product and the system figured out what it is, suggested a sensible selling price, and updated stock on the spot? No manual lookups, no guessing.
Making it real meant integrating the Gemini SDK into a Next.js app backed by PostgreSQL and Supabase. The AI identifies the product, proposes pricing, and even surfaces simple business analysis — turning a barcode scanner into a small business assistant.
I wore two hats on this one: Fullstack Developer and Project Manager. The technical challenges were fun, but the bigger lesson was product thinking. Features only matter if they survive contact with a real cashier on a busy afternoon.
Iscanow taught me that AI isn't magic — it's a tool that becomes powerful when you wrap it in a thoughtful workflow. That belief now drives everything I do as an AI Engineer.