Fuel
Calorie tracking that doesn't feel like data entry.
Type what you ate. AI handles the rest.
Three steps.
Under 10 seconds.
Type what you ate
Plain English. "2 eggs, sourdough toast, black coffee." No searching through databases. No measuring cups.
AI parses it
An LLM breaks down your text into individual foods with calories, protein, carbs, and fat. Confidence scores let you know when to double-check.
Confirm and log
Review the parsed items, adjust if needed, and log. Your macros, ring chart, and daily totals update instantly.
Just describe
your food.
No scrolling through databases. No scanning nutrition labels. Type "leftover pad thai, maybe half a serving" and the AI handles the ambiguity, estimates portions, and returns structured macro data with confidence scores.
For packaged foods, point your camera at the barcode. Between AI and barcode scanning, you can log a full day of eating in under 60 seconds.
Everything you need.
Nothing you don't.
AI Food Parsing
Type in plain English. The LLM returns structured nutrition data with calories, macros, and confidence scores.
Barcode Scanning
Point your camera at any UPC. Vision framework reads the barcode and pulls nutrition data instantly.
HealthKit Sync
Calories, macros, and weight sync directly to Apple Health. Your data flows where you need it.
Apple Watch
Quick-log meals from your wrist. See remaining calories and macros without pulling out your phone.
Home Screen Widget
WidgetKit-powered glanceable view. Calories remaining and macro rings right on your home screen.
No Account Required
All data stays on your device. No sign-up, no cloud sync required. The only network call is to the AI parsing API.
Native iOS.
No compromises.
How it all connects.
User Input (text / barcode / template)
|
+--------+--------+
| |
LLM API Vision Framework
(NLP parse) (barcode scan)
| |
+--------+--------+
|
SwiftData
(FoodEntry model)
|
+---------+---------+
| | |
Dashboard HealthKit Watch App