Bedrock Managed Knowledge Base adds ServiceNow connector and self-serve setup

These updates lower the operational friction of enterprise RAG. The native ServiceNow connector pulls knowledge articles and service-catalog items directly into a fully managed knowledge base, letting agents answer IT and operations questions grounded in an organization's own ServiceNow data without custom ingestion pipelines. The addition targets one of the most common enterprise knowledge stores.

The self-serve (3LO, three-legged OAuth) setup for SharePoint, OneDrive and Confluence is a governance win: previously connecting these sources required admin-level 2LO credentials, a bottleneck that put IT in the loop for every data source. Now individual users can authorize their own accessible content, speeding deployment while respecting existing permissions. Automatic sync scheduling completes the package — users no longer manually trigger a re-sync when source data changes, keeping the knowledge base fresh without managing pipelines.

Taken together, the trio reflects AWS's strategy of making Bedrock's managed RAG a low-effort default for enterprises rather than a build-it-yourself framework, competing with Microsoft's Foundry knowledge features and Google's grounding tools. The broader context is a week where AWS shipped across the agentic stack — AgentCore memory, MCP federated data access, the WhatsApp assistant — signaling Bedrock as an end-to-end enterprise AI platform. The main caveat is connector maturity: native connectors simplify setup but can lag on edge cases (permission mapping, incremental crawl fidelity) that enterprises hit at scale. Watch for how the 3LO model handles permission revocation and whether more connectors (Salesforce, Jira) follow.