AWS announced the availability of Amazon CloudWatch Omni, a unified experience for application monitoring and AI agents that organizes log, metric, trace, and alert data around an application and its relationships instead of displaying them as separate signals. The service provides a shared workspace that engineers can use through a dedicated organization URL and enterprise sign-in, without requiring access to the AWS Management Console.
CloudWatch Omni is based on OpenTelemetry, so telemetry data that organizations send to CloudWatch appears in the new experience without reconfiguration. Other workloads instrumented with OpenTelemetry can also send their data to an OpenTelemetry Protocol (OTLP) endpoint.
What does the service offer?
Omni automatically discovers services and builds a map of their dependencies, based on telemetry data and resource discovery through AWS Config. Instead of relying on static dashboards that require ongoing maintenance, teams can define what they consider important, such as availability targets, latency budgets, and error-rate thresholds, allowing monitoring views to adapt as the application changes and new services are added.
The service provides each team with what AWS calls a Space containing the applications it owns and the data associated with them. The space references existing CloudWatch data without additional data movement. Users can also ask natural-language questions about their applications, analyze telemetry data, and display indicators related to system health.
Shared investigations with DevOps Agent
Amazon DevOps Agent participates in investigation sessions alongside engineers, relying on the same telemetry data the team sees. AWS says the agent connects events across services, traces root-cause paths through the dependency graph, and suggests next steps, while retaining an investigation record for use in subsequent incident reviews.
In the scenario presented by AWS, a rise in the error rate of the payment service opens a session that includes the application topology, a recent deployment, and increased latency from an external payment interface, along with the agent’s initial analysis. An engineer from the payments team can join the same session and see the accumulated context before identifying the root cause and rolling back the change.
What changes in practice?
The primary change is not replacing existing CloudWatch tools, but adding an application-centric interface and collaboration during investigations. Existing alerts, dashboards, APIs, and console operations continue without change. Engineers also do not need AWS Console permissions to use Omni after setting up enterprise sign-in through IAM Identity Center, which supports Okta, Azure AD, and other SAML 2.0 providers.
Existing CloudWatch customers can start from the “Try CloudWatch Omni” option in the CloudWatch console, then configure the identity provider and create spaces for teams and environments. AWS provides connectors for bringing telemetry data from other environments into the same spaces and investigation sessions. According to the material, DevOps Agent is enabled by default in every investigation session within Omni.
Availability and limitations
CloudWatch Omni is now available, while AWS refers users to the Amazon CloudWatch pricing page for pricing details. The material does not provide pricing figures or a list of regional availability. The fact that investigations depend on available telemetry data and enterprise identity configuration also means that the value of the experience will remain tied to the quality of each organization’s telemetry configuration and coverage, a point teams should assess before expanding usage.