PulseBrew is a smart, offline-first, AI-driven coffee companion app that monitors user health signals and provides intelligent caffeine reminders.

Architecture: Flutter (Android) + FastAPI (Railway) + Supabase (Sync) + SQFlite (Offline).

Backend: https://pulsebrew-production.up.railway.app/

• 📊 Health Data Integration
  • Reads heart rate via health package (with Mock fallback for emulators).
• 🧠 Deterministic AI
  • Uses CaffeineDecayModel (Exponential decay, 5h half-life) to predict plasma levels.
• 📶 Offline-First
  • Works 100% offline using sqflite.
  • Syncs to Supabase when online via SyncRepository.
• ⚡ FastAPI Backend
  • Validates logic and provides backup recommendations.

• Mobile: Flutter 3.x, Provider, Sqflite, Supabase Flutter.
• Backend: FastAPI, Python 3.10+, Pydantic.
• Database: Supabase (PostgreSQL), SQLite (Local).
• Infrastructure: Railway (Free Tier).

pulsebrew/
├── backend/            # FastAPI Backend
│   ├── main.py
│   ├── logic.py        # AI Logic
│   ├── models.py       # Pydantic Models
│   └── schema.sql      # Database Schema
├── pulsebrew_app/      # Flutter App
│   ├── lib/
│   │   ├── services/   # LocalDB, Sync, Biometrics
│   │   ├── logic/      # AI Logic (Dart)
│   │   └── ...
│   ├── android/
│   └── pubspec.yaml
└── README.md

cd backend
pip install -r requirements.txt
uvicorn main:app --reload

cd pulsebrew_app
flutter pub get
flutter run

MIT License

Mohith Anjan
Refactored by Antigravity