Retrieval-Augmented Generation(RAG)
An AI architecture that retrieves relevant documents from a knowledge base and feeds them to a model so the model answers from your data.
In detail
Retrieval-Augmented Generation (RAG) combines a search step (retrieval) with a generation step. Your documents are converted to vector embeddings and stored in a vector database. When a user asks a question, the system retrieves the most relevant chunks and feeds them to a language model along with the question, so the model answers using your content rather than its general training data. RAG is the dominant pattern for AI assistants that need to know about a specific organisation's policies, products or knowledge base.
Why it matters for Australian business
For an Australian business RAG is how you build an internal Q&A assistant or customer-facing helper that knows your policies, contracts or product documentation, without needing to retrain a model. Cost is usually one to two orders of magnitude lower than fine-tuning. The risks are that retrieval can miss relevant chunks (poor recall) and that sensitive content gets into prompts that flow to a third-party model. Australian Privacy Act considerations apply to the retrieval pipeline if it touches personal data.