RAG & Retrieval · Ground answers in evidence
What is retrieval-augmented generation?
Short answer
RAG retrieves relevant source passages at query time and adds them to the model's context so the answer is grounded in evidence and can cite where it came from.
Retrieval-augmented generation separates knowledge from the model. Instead of relying on what the model memorised during training, the system searches an external corpus for relevant passages and places them in the prompt. The model then reads, synthesises and cites that evidence.
- The corpus stays current without retraining the model.
- Permissions and freshness can be enforced at retrieval time.
- Answers can cite sources, which makes verification possible.
- Retrieval quality becomes a first-class part of product quality.
Follow-ups you should expect
Interviewers use these to check whether the first answer was understanding or memorisation.
- 1What kinds of questions is RAG a poor fit for?
- 2How is RAG different from search?
- 3What would you measure to know RAG is working?
Say it in the interview
RAG is a way of grounding generation in external knowledge. At query time we retrieve the passages most likely to contain the answer, put them in the prompt, and ask the model to answer from them with citations. The advantage is that knowledge can change without retraining, and access can be permission-aware. The caveat is that the answer is only as good as the retrieval, so I measure retrieval and generation separately.
Use this as a model, not a script. Saying it in your own words is what makes it survive follow-up pressure.
Related questions
Reading is step one
Now say it out loud.
Understanding an answer and delivering it under pressure are different skills. Rehearse this topic in a live RAG & Retrieval mock interview and get scored feedback on depth, clarity and confidence.