These are RAG questions of the kind Microsoft actually asks — the patterns reported from Microsoft's AI-engineering rounds, where design trade-offs, scale and failure modes matter as much as definitions. Treat this page as a mock interview: say every answer out loud before revealing it. If one surprises you, the lesson behind it is linked at the bottom.
Microsoft RAG concept questions
Your RAG assistant serves multiple enterprise tenants from one deployment. How do you guarantee no tenant can ever retrieve another tenant's documents?
Asked in
Recall@5 for your system is 92%, but users still report bad answers. What are the likely causes?
Asked in
Where does prompt injection fit into a RAG system, and what makes RAG specifically vulnerable?
Asked in
Microsoft RAG applied & hands-on questions
Write the retrieval call for a multi-tenant assistant, and list the three places outside retrieval where the same permission scope must also be applied.
Asked in
How to use this page: Microsoft rarely asks something you've never seen — they ask a standard RAG concept and then push one level deeper ("why?", "what would you do if..."). Master the concept in the RAG course lessons, and the follow-up stops being scary.
Keep practising: What Is RAG?, Chunking, Retrieval Techniques and Evaluating RAG cover what most Microsoft RAG rounds test.