Project - Yammmi – AI Calorie Tracker
With advanced AI, losing weight and reaching your goal weight is now as simple as taking a picture.
- Client
- Own Project
- Year
- Service
- App development, Backend, API, AI
The Challenge — Manual Tracking ≠ Sustainable
Staying in a calorie deficit is still the most reliable way to lose fat, yet traditional food journals and barcode scanners demand time, scales, and sheer willpower—especially when you're eating out. Most people quit within two weeks, long before any progress is visible.
Many people think the gym is where fat loss happens. Training matters, but lasting results come from what — and how much — lands on your plate. Real change starts with nutrition, and that’s exactly where Yammmi delivers.
Our Approach — 1 Photo = Full Nutrition
Take a quick shot of your plate and Yammmi’s computer-vision model
outputs calories, macros, and 30+ micronutrients in under 20 seconds—no
weighing, no guessing. The model has been trained on
339 k+ labelled dishes (and counting) for 96 % recognition accuracy.
Top tip
A single photo can’t capture every ingredient or cooking method.
If your dish is complex, just add a few keywords — "extra cheese", "grilled in butter", "almond topping" — and Yammmi's AI will factor them into the analysis for even greater accuracy.
Making Progress Visible — From Plate to Goal Weight
Yammmi does not stop at a single food photo. The app connects meals, daily targets, activity, and weight history into one understandable view, so progress is not just estimated, but visible.
Download Yammmi today
What we did
- Idea
- Planning, Concept & Design
- App Development
- Database & Server
- APIs & Integrations
- AI Integration
- Prompt Engineering
- Continuous Improvement
Yammmi started as a passion project. I’d tried countless calorie-tracking apps to keep my diet aligned with my workouts, but the constant weighing and tedious logging always made me give up. That frustration sparked the idea for Yammmi: a faster, simpler way to track meals so you can stay on track for the long haul.
CEO of Wapps GmbH
- Per meal
- 1 photo
- To log
- < 20 s
- Accuracy
- 96 %
- Avg. in 6 weeks
- –3.1 kg