StillGood
Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote
StillGood helps families use the food they already bought before it goes bad. The average US family of four throws away $2,913 of food a year — over 11% of what they spend on food (U.S. EPA, 2025) — mostly because nobody can keep track of what's in the fridge. StillGood removes the typing. Snap a grocery receipt or a photo of your fridge, and Meta's Muse Spark reads every item — even store abbreviations like "GRN ONION BNCH" — and estimates when each one will go off. From one photo of a real family fridge, it found 17 items in about 20 seconds. The app shows what to use first, sends a daily "use soon" reminder, and plans dinners around the food closest to its date. Every suggestion fits the household: favorite cuisines, cooking time and skill, and each person's allergies and dietary restrictions, which are treated as strict rules. Recipes can be adjusted in plain words, and the shopping list only shows what's missing. You can also just tell the assistant "we finished the milk" — it always shows the change and asks before saving. A savings screen tracks money rescued and waste avoided, and a built-in before-and-after survey measures real impact. Built with Next.js and Muse Spark (vision, structured output and reasoning), all AI calls server-side. Plus is $5.99/month with a free starter pack.