Amazon Bedrock Managed Knowledge Base: What Changes for Developers
AWS's Amazon Bedrock Managed Knowledge Base automates RAG ingestion, parsing, and retrieval — key quota limits and userContext security to know.
AWS launched Amazon Bedrock Managed Knowledge Base on June 17, 2026, a fully managed RAG service that handles ingestion, parsing, embeddings, vector storage, hybrid retrieval, and re-ranking for enterprise sources like S3, SharePoint, Confluence, Google Drive, and OneDrive.
Smart Parsing automatically routes each file type through Bedrock Data Automation or Nova Multimodal Embeddings, replacing hand-built Textract and Transcribe pipelines — though custom connectors still earn their place for internal platforms or unusual permission models.
Retrieval splits into three modes: standard hybrid Retrieve for simple lookups, managed re-ranking to cut noisy chunks, and AgenticRetrieveStream for multi-document questions requiring iterative decomposition. Strict per-minute quotas, including a 300-request service limit, make query routing and backoff logic essential design considerations.
Integration with Bedrock AgentCore Gateway exposes retrieval as MCP-compatible tools for frameworks like LangChain and CrewAI, but document-level security still requires applications to pass userContext explicitly — IAM identity alone won't enforce per-user filtering.