Build production RAG systems with semantic chunking, incremental indexing, and filtered retrieval.Use when implementing document ingestion pipelines, vector search with Qdrant, or context-awareretrieval. Covers chunking strategies, change detection, payload indexing, and context expansion.NOT when doing simple similarity search without production requirements.