Back to all projects
Featured
MacroTrackr logs your meals for you. Point your camera or describe a meal in plain English and the AI estimates calories and macros — no database searching. Daily dashboards, goal tracking and history analytics keep the friction at zero.
Stack & tags:
Next.js 16React 19TypeScriptTailwind v4SupabasePostgresVercel
Launch MacroTrackr
Production build · live at macrotrackr.app · no install required.
Recent announcements & updatesSee all
v1.55.0
15 Sep 2026
Every feature page row shows a static slice of the app screen it describes (confidence chips, portion fix, calorie ring and macro rows, week bars, trend line, weekday averages, activity heatmap, product sheet with Nutri-Score/NOVA, search, saved meal, pinned rail), built from the app tokens, instead of 15 hatched placeholder boxes with fake 01/02/03 counters.
v1.54.1
14 Sep 2026
Numbers followed the device locale instead of the page's and kept stray decimals: an English page on a French machine read "2 406,9 kcal / day" next to "69.2 g", and French pages read "86.5 kg". Calories, grams, weights and paces now round and group with the page locale across Goals, Trends, Today, History, the meal sheet and onboarding; the goal progress bar shows the full quantity in the user's units instead of raw kilograms.
About this project
MacroTrackr replaces manual food logging with a single AI step: snap a photo or type what you ate, and it returns an estimate of calories and macros. The goal is to remove the friction that makes most tracking apps get abandoned in a week.
Behind the screen, meals are parsed by an AI vision + natural-language layer, then persisted to Postgres through Supabase with auth and row-level security. The front end is server-rendered with Next.js for fast first paint and live dashboards.
What's inside:
- Photo and natural-language meal logging
- AI calorie + macro estimation
- Daily dashboards and goal tracking
- History analytics over time