Concepts
RAG Chain
Combines a vector store with an existing chain — retrieves relevant context, injects it into the prompt, generates a grounded answer.
Why it exists
Lets a chain answer questions using information it wasn't trained on, by retrieving relevant text first and feeding it into the prompt before generation.
educhain.py
from educhain.core.rag import RAGChain
rag = RAGChain(vector_store=store, chain=chain)