You are developing a Retrieval-Augmented Generation (RAG) solution for a company.

I response expiry: set a 24-hour TTL on each key. Embedding freshness: delete related cache keys when the source document changes.

Detailed Explanation: A fixed TTL encodes the requirement that an AI response must expire after a defined period rather than survive indefinitely or have its lifetime extended on every read. Embedding freshness is a separate consistency requirement: when the authoritative document changes, any cached representation derived from the old document must be invalidated immediately. Memory eviction policies are capacity-management mechanisms and cannot guarantee a precise 24-hour lifetime or source-data consistency.

Study Guide Alignment: AI data-management workloads: Cosmos DB, PostgreSQL, caching, vector storage, vector retrieval, consistency, and connection optimization.

Official Microsoft Learn References: AI-200 Study Guide | Azure Managed Redis documentation