AWS made the next generation of Resilience Hub generally available on May 28. The release combines a new application model with dependency discovery, AI-assisted failure-mode analysis, and organization-wide reporting.
Model the customer outcome
The announcement organizes applications into systems, user journeys, and services. It also describes discovery of AWS services and internal or third-party endpoints, alongside resilience policies managed across an organization. Existing customers can keep their current experience while planning adoption.
For SRE teams, the useful change is an opportunity to connect infrastructure findings to a specific customer action. A dependency that looks minor in a resource inventory may be critical to completing checkout, signing in, or submitting a job.
Our analysis: this model needs an accountable service owner. A generated dependency view will be most useful when someone can confirm which paths are required, which have fallbacks, and which are obsolete.
Turn recommendations into experiments
Choose one important journey and compare its assessment with the team’s current incident history and recovery runbook. Investigate disagreements instead of treating either source as automatically correct.
Then select a failure scenario that can be exercised safely. Define the expected customer impact, a stop condition, and evidence that recovery actually completed. Record any dependency the assessment missed or any recommended action that does not match the application.
An AI-generated assessment is an input to that work, not proof that a workload can survive an outage. Keep the distinction visible in reporting: a discovered risk, an accepted recommendation, and a verified recovery result are different states. The best initial rollout connects those states for one journey before expanding the dashboard across an organization.
See the original announcement for availability and release details.