How dadiya-harsh/cure-now works
Project type and purpose
• This is a full-stack web application (“Cure-Now Health Chatbot”) with a Python/Flask backend and a static HTML/JS/CSS frontend.
• Its purpose is to serve as a medical chatbot: it takes user queries, issues vector-search lookups against a medical knowledge base, calls Google’s Gemini AI for answer generation, and returns responses in real time.
File structure
• Top-level:
• alembic/ → database migration scripts (SQLAlchemy + Alembic)
• backend/ → Flask application
• research/ → Jupyter notebook for experiments
• requirements.txt, README.md, .env template
• backend/app/ subfolders:
• routes/ → Flask blueprints (auth_routes.py, chat_routes.py, session_routes.py)
• models.py + config.py → SQLAlchemy models and configuration
• utils/ → embedding generation (Hugging Face), Pinecone ops, Gemini AI ops
• static/ + templates/ → frontend assets and Jinja2 templates
• backend/run.py → Flask application entrypoint
Review README insights
• Tech stack: Python/Flask, LangChain, Hugging Face embeddings, Pinecone vector DB, Google Gemini AI, HTML/CSS/JS frontend
• Features & flow: user → chat UI → Flask endpoints → utils (vector search + AI) → external services → UI