Metoro

The tool deploys an AI‑driven SRE agent to Kubernetes clusters, collecting logs, metrics, traces, profiling data, Kubernetes events, resources, and deployment context through an eBPF‑based collector that runs as a DaemonSet. Because collection occurs in the kernel, no SDKs, code changes, or container restarts are required, and the data is pre‑correlated with Kubernetes identities for immediate use.

Using the gathered telemetry, the AI agent automatically detects production issues, investigates alerts, verifies each deployment, identifies root causes, and can generate remediation pull‑requests. The system supports a unified query language across all signals and is compatible with OpenTelemetry and Prometheus formats.

Installation is performed via a single Helm chart, bringing the agent online in under five minutes on any major Kubernetes platform (EKS, GKE, AKS, OpenShift, etc.). The solution is positioned as an experimental utility for teams that want continuous, automated observability and self‑healing capabilities without modifying application code.