What is this project?

Food logging fails because it is tedious. NutriScan removes the search-and-select step entirely: type "I had 2 scrambled eggs cooked in butter, 2 slices of brown toast with peanut butter, and a glass of whole milk" and it returns a full itemised breakdown.

Each detected food becomes a row with its estimated quantity, calories, protein, carbs and fat — scrambled eggs at 140 kcal, butter at 102, brown toast at 160, peanut butter at 188, whole milk at 150. The summary rail totals it (740 kcal) and shows each macro both in grams and as a share of total calories, so the balance of a meal is visible at a glance rather than buried in numbers.

The app is honest about its estimates — a standing note reminds users that portions are AI-estimated and that naming the cooking method improves fat accuracy. Analysed meals can be logged, revisited from History, saved to Favorites, or rolled up in Insights.