What it is

Retrieval-augmented generation (RAG) is a technique where an AI looks up real information before answering your question, rather than relying purely on what it memorised during training. It's like the difference between answering an exam from memory versus being allowed to check your notes. The AI retrieves relevant documents, feeds them into its context, and generates an answer based on actual sources. This dramatically reduces the made-up-nonsense problem.

Why it matters for your job

RAG is how companies are making AI actually useful for work, not just general chitchat. When an AI can pull from your company's internal documents, policies, and data before responding, it goes from a clever toy to a practical tool. If you understand how RAG works, you can help your team set it up properly, choose what documents to include, and spot when it's pulling from outdated sources. That's a role that didn't exist two years ago.

What to do about it

If your company is implementing AI tools, ask whether they use RAG. If they don't, suggest it. If they do, get involved in deciding what information the AI has access to. Your domain knowledge about which documents matter is exactly what makes RAG work well.