Bedrock adds scheduled syncs for native knowledge-base connectors

Amazon Web Services announced automatic sync scheduling for Amazon Bedrock Managed Knowledge Base on September 4, 2026. Engineers can configure daily, weekly or monthly syncs for all native data source connectors, according to the AWS announcement.

Availability and supported schedules

AWS describes the feature as available now. The announcement does not specify regional availability or explicitly designate the release as preview or general availability.

The supported cadences provide a managed alternative to manually triggering syncs or maintaining a custom scheduler. AWS gives examples of daily updates for Confluence support content, weekly updates for SharePoint policy documents and monthly updates for reference material in Amazon S3. These are illustrative workloads, not prescribed schedules.

Which ingestion jobs could go away

For a retrieval-augmented generation application whose ingestion job primarily invokes a connector sync on a schedule, the new setting could replace that scheduling layer. AWS says this reduces operational overhead; the practical benefit depends on how much work the existing job performs beyond triggering ingestion.

A scheduler that also validates documents, coordinates upstream exports or enforces application-specific checks is a different case. The announcement does not establish replacements for those steps. Teams should separate trigger logic from other ingestion responsibilities before removing a job.

Freshness still needs an explicit target

Scheduled sync does not mean continuous or immediate ingestion. Content changed just after a scheduled run may wait until the next run, plus processing time, before becoming retrievable.

Teams should choose a cadence against their freshness requirements and verify how failures are surfaced and recovered. The announcement does not document retry behavior or completion-time guarantees, so scheduling alone is not enough to establish a freshness commitment.