Habit Tracker
Habit Tracker
Section titled “Habit Tracker”System Overview
Section titled “System Overview”Build a local-first habit tracker focused on bottle pickup tracking, habit logging, dashboard analytics, and later camera-based event detection.
Main parts
Section titled “Main parts”- Frontend: React app for dashboard, logging, and chat
- Backend: FastAPI for APIs, business logic, and AI endpoints
- Database: Postgres as source of truth
- Vector store: pgvector inside Postgres for embeddings
- AI orchestration: LangGraph for multi-step AI flows
- Future vision service: separate local service for camera or YOLO-based event detection
Product direction
Section titled “Product direction”The app is product-first, not RAG-first.
The main data is structured application data:
- habits
- habit logs
- bottle events
- notes
Later, AI and vision features will build on top of this data.
Initial MVP
Section titled “Initial MVP”- manual habit logging
- bottle pickup logging
- dashboard summaries
- basic AI chat over structured data
Future scope
Section titled “Future scope”- semantic retrieval over notes
- thread-based AI assistant
- vision events from camera
- posture tracking