This gives Claude persistent memory across conversations by storing information with semantic search and knowledge graph capabilities. You'd reach for it when you need an AI assistant to remember context, relationships, and facts beyond single sessions. Think customer details, project knowledge, or personal preferences that should persist and be retrievable through meaning rather than exact keywords. The knowledge graph layer means it can track how pieces of information relate to each other, not just store isolated facts. Useful for building assistants that need to accumulate context over time and recall relevant information intelligently.