RAG, Agents, and Compound AI Systems

5.34.1Retrieval, Reranking, Generation, and the Index Pipeline#

5.34.2Vector Databases, Embeddings, and Retrieval Caching#

5.34.3Tool Calling, Sandboxing, and Environment Resource Management#

5.34.4Multi-Turn Sessions, Long-Term State, and Context Management#

5.34.5Agent Workflows, Parallel Tools, and Long-Tail Task Scheduling#

5.34.6Model Routing, Cascades, and System-Level Caching#

5.34.7End-to-End Quality, Latency, and Cost of Compound Tasks#