DEEP EXPLANATION

CRAG and Fallback Web Search Patterns (EXPLAINED)

Scenario BasedRAGHard16 min read

Hard RAG pattern question on corrective retrieval and gated web fallbacks.

TL;DR — Quick Answer

CRAG evaluates retrieved evidence quality and triggers corrective actions (rewrite query, re-retrieve, or web search) when confidence is low. Gate web fallback behind low local scores, require citations from fetched pages, and still abstain if evidence is weak — do not freely mix untrusted web text into high-risk answers.

The Interview Question

Explain Corrective RAG (CRAG). How would you add a web-search fallback when local retrieval confidence is low without increasing hallucination risk?

Deep Explanation

Get deep explanations, PDF export & all RAG questions

4 more sections of deep explanation
Real-world examples
Common mistakes
Interviewer expectations
Follow-up questions

CRAGRAGWeb SearchFallbackOpenAIGooglePerplexity

Up next

Next Question

Self-RAG and Corrective RAG (EXPLAINED)

Hard RAG interview question on self-rag and corrective rag — architecture, trade-offs, eval, and production patterns.

Continue