AI

RAG

Retrieval-Augmented Generation — feeding an LLM relevant documents at query time so it answers from your data.

RAG patches the LLM's two biggest weaknesses — it doesn't know your private data and it makes things up — by retrieving relevant facts and injecting them into the prompt before the model answers. The flow: embed the question, search a vector database for matching chunks, hand those chunks to the LLM as context, get a grounded answer. It's how you build a chatbot over your own docs without retraining a model. For most 'answer from my knowledge base' cases it's cheaper, faster, and easier to update than fine-tuning.