Inspiration
Outbreaks happen in your food. This can be troubling for the elderly, disabled, and vulnerable individuals. We make sure the damage never happens.
What it does
After providing your religious and personal preferences, medical conditions, live health data, and current diet, we're able to diagnose if there's any major news issues with the product you're looking at, and recommend the better alternative whilst looking out for your health.
How we built it
- Framework: Flutter / Dart (Material 3)
- Backend: Firebase (Auth anonymous sign-in, Firestore, hosting)
- APIs: FoodBase (product photo ID), Veryfi (receipt OCR), USDA FoodData Central (nutrients), Tesco API, OpenRouter (LLM chat + tool calling)
- On-device ML: ML Kit text recognition (OCR fallback), Google ML Kit camera analysis
- Voice: speech_to_text, flutter_tts, on-device Kokoro TTS (sherpa_onnx + piper_phonemize), audioplayers
- Health: Health Connect (native Android channel — HR, steps, sleep, BP)
- Camera: camera plugin for product/receipt capture
Challenges we ran into
Decomposing the project and architecting the solution to encompass the right data paths to provide the most optimal output.
Accomplishments that we're proud of
A working solution.
What we learned
Many new technologies.
What's next for camelCase
Including exercise data.
