Amazon CloudWatch Omni: AI-first observability for agents and applications
AWS introduces Amazon CloudWatch Omni, an AI-driven observability tool unifying application and agent monitoring across clouds and IDEs with natural language interaction and dependency mapping.
Amazon CloudWatch Omni, now generally available, expands AWS’s observability suite by integrating AI-powered analysis with OpenTelemetry standards and CloudWatch’s existing infrastructure. The service centralizes telemetry from AWS accounts, regions, and external clouds such as Azure, allowing teams to monitor workloads in a single interface. Users can access Omni via a standalone web console with single sign-on or through IDE extensions for local development, supporting tools like VS Code and Cursor.
Omni automates service discovery and dependency mapping, presenting key metrics to streamline troubleshooting and reduce manual investigation time. Operators can query telemetry using natural language, generate dynamic visualizations, and trace issues to root causes with assistance from AWS DevOps Agent. The console also offers guided point-and-click workflows for those preferring manual exploration of signals, traces, or evaluations.
A dedicated agent observability mode supports AI workloads across frameworks including LangGraph, CrewAI, and OpenAI Agents SDK, enabling evaluation-driven development. For each prompt, model call, and tool invocation, Omni provides quality assessments and experiment validation to ensure fixes before deployment. This workflow targets developers building agentic systems who need reproducible testing and debugging.
To begin using Omni, teams create a space in the CloudWatch console, configure SSO, and log into the web experience. Agent developers can install the free Omni extension in supported IDEs to instrument, debug, and evaluate agents locally without an AWS account. Omni is available in US East (N. Virginia), US West (Oregon), and Europe (Ireland) regions.