About FirstAid-LLM
What is FirstAid-LLM?
FirstAid-LLM is an AI-powered first-aid guidance system that provides instant, evidence-based answers to first-aid questions. It uses Retrieval-Augmented Generation (RAG) to search through 99+ authoritative medical guidelines and compose clear, citation-backed responses.
How It Works
- You ask a question about a first-aid scenario
- AI searches our vector database of 5,900+ indexed chunks
- Relevant sources are retrieved using semantic similarity
- Gemini 2.0 Flash composes a clear answer with citations
- You receive guidance backed by authoritative sources
Technology Stack
- Frontend: Flask + Jinja2 + Vanilla JavaScript
- Backend: Python 3.12
- Vector Database: ChromaDB
- LLM: Google Gemini 2.0 Flash (via Vertex AI)
- Embeddings: text-embedding-004
- Deployment: Docker + Docker Compose
Academic Project
This project was developed for AC215 - Advanced Practical Data Science at Harvard University (Fall 2025). It demonstrates:
- RAG system design and implementation
- Vector database integration
- LLM prompt engineering
- Full-stack web development
- Docker containerization
- CI/CD pipelines
⚠️ Important Limitations
FirstAid-LLM is:
- ✅ Good for: Learning basic first-aid procedures
- ✅ Good for: Quick reference in non-emergency situations
- ❌ NOT a replacement for: Professional medical advice
- ❌ NOT for: Medical diagnosis or treatment decisions
- ❌ NOT for: Emergency situations (call 911 immediately)