RAG and retrieval
RAG vs Fine-tune vs Long-Context interview questions
The rag vs fine-tune vs long-context questions that keep coming up in AI engineer interviews. The first 3 come with full answers.
Updated 2026-09-05
01
What are the three ways to give a model knowledge, and the one-line pitch for each?
Three tools for three jobs. RAG (retrieval-augmented generation) keeps facts outside the model and fetches the relevant ones at question time — look it up. Fine-tuning nudges the model's weights so it behaves a certain way — bake it in. Long context simply drops all the needed text into the prompt for one request — paste it in. The mistake interviewers listen for is treating them as rivals: they solve different problems and combine well.
When is RAG the right choice — what properties of the knowledge push you toward it?
Reach for RAG whenever the knowledge is large, private, or changing. It keeps the facts outside the model in a vector store, so editing a document makes the system current the same day — no retraining — and it can cite which source it used. That makes it the natural fit for Q&A over your own docs, a help centre, or any knowledge base that keeps moving. The alternative, baking facts into weights, is stale the moment a doc changes.
Why is fine-tuning for behaviour rather than facts, and how do you pair it with RAG?
Fine-tuning shows the model many examples and adjusts its weights to match a behaviour — a style, tone, strict format, or narrow skill. It changes how it responds, not what it knows, so cramming changing facts in is expensive and goes stale. The clean pairing: fine-tune the behaviour, retrieve the facts with RAG. A support bot might be a lightly fine-tuned model (brand voice, tidy format) that pulls current help articles per question via RAG.
Also asked in interviews
- When does long context beat both RAG and fine-tuning, and what costs and failure mode limit it?
- How would you actually decide between them for a new feature?
- Using the open-book / years-of-training / whole-binder analogy, map each phrase to RAG, fine-tuning, and long context.
These 3 run inside the bootcamp as recall drills and voice mock interviews. You answer, the AI grades.
Knowing an answer and saying it under pressure are two different days. Inside Skillumen you answer these out loud and get graded on the spot. Foundations is free.
Try it free →
The open-source companion list, Awesome AI Engineer Interview Questions, curates 105 of these on GitHub.