← Back to all projects
🥗

My product Open source

FitTrack-AI

Nutrition from a photo of your meal

Take a photo of a meal and get its nutrition — with values grounded in real food databases rather than model guesses.

vision LLMsUSDA dataOpen Food Facts

Built for: Anyone tracking what they eat.

↓ Scroll to explore the problem, architecture and stack

😩

The problem

Logging food by hand is tedious, and AI apps that "guess" calories from a photo are often confidently wrong.

💡

Why it matters

Recognising the food with a vision model but taking the numbers from USDA and Open Food Facts makes the results quick and trustworthy.

What you can do with it

01Snap a photo of a meal.
02See recognised foods and their nutrition values.
03Trust numbers that come from real food databases.

How it’s built

A simplified view of the architecture — follow the dots as a request moves down the layers.

You
Next.js appupload a photo
API
FastAPI
Intelligence
Vision LLMrecognise foods
Ground truth
USDA FoodData Central
Open Food Facts

A request, step by step

  1. You upload a photo.
  2. A vision model identifies the foods.
  3. Nutrition values are looked up in USDA and Open Food Facts, not guessed.

Tech stack

Backend
PythonFastAPI
AI
Vision LLMs
Data
USDA FoodData CentralOpen Food Facts
Frontend
Next.js

What I built

  • Meal recognition from a photo using vision LLMs.
  • Nutrition values grounded in USDA FoodData Central and Open Food Facts.
  • A Next.js frontend on a FastAPI backend.