Step 3 of 5 — Query Knowledge Base
RAG Engine
Retrieval-Augmented Generation — ask natural language questions and get AI answers backed by real hospital documents.
Pipeline Architecture
⌨
Query Input
Natural language question
→
🧮
Embed Query
Text → vector representation
→
🗄
ChromaDB Search
Cosine similarity search
→
📑
Top-K Chunks
Most relevant document snippets
→
🤖
LLM Synthesis
GPT generates a grounded answer
Ask questions about healthcare gaps
Tech Stack
LlamaIndexChromaDBOpenAI EmbeddingsGPT-4o-miniFastAPINext.js