Amazon Bedrock Managed Knowledge Base adds single-shot RAG and retrieval-only modes to the same API as its multi-step agentic retrieval
Search once for simple questions, several times for hard ones, without two systems. How each of the three modes is meant to be used.
// Key points
- The new modes come through AgenticRetrieveStream: single-shot RAG searches once and has a foundation model answer from the passages it finds, while retrieval-only returns the matching passages without an answer.
- The existing agentic retrieval breaks a question into sub-queries, checks whether the results are enough and, if not, searches again within a configured iteration limit, retrieving a full document when a passage lacks context.
- AWS's example is an internal banking policy assistant: single-shot RAG for an account-opening checklist, agentic retrieval for comparing policies across business types.
Builder's takeMatching retrieval effort to question difficulty is something I have wanted in the AI Cloud Drive's document Q&A. You can copy the structure off AWS too: a cheap classifier decides whether to go multi-step, and a separate source-only entry point lets users check passages without waiting for generation.
// Background · from #AI agents
Full timeline →- Nov 10 AWS launches Nitro Sandboxes: each EKS pod, ECS task or lightweight VM gets its own hardware-isolated VM, enabled with a one-line change at no additional charge
- Nov 9 Google measures agent distraction: 13% of sessions meet bug-referencing TODOs, and detoured sessions take 70% more steps and 77% more tokens
- Nov 9 Greenlight builds Luna, a security operations agent on AgentCore: up to 4,000 cases a month, more than 90% resolved without human intervention