Retrieval-augmented generation (RAG)
Retrieval-augmented generation, or RAG, is a technique where an AI first looks up relevant source text, then writes its answer from that text.
A plain language model answers from patterns in its training data, which it cannot cite and can get wrong. RAG adds a retrieval step. It searches a trusted set of documents, pulls the passages that match your question, and grounds the answer in them.
This is what makes citations and fewer made-up answers possible. The model is steered by real source text instead of guessing from memory.
How StudyPDF does this
Bo uses retrieval over your own uploaded course, so it answers from the right passages in your material and can show you the page each one came from.
Get started free