Workflow Examples

These diagrams illustrate concrete scenarios of how SmartSRE agents interact with your infrastructure, handle risks, and manage lifecycles.

Scenario 1: Autonomous Scaling (Happy Path)

In this scenario, a Cloud Run service experiences a CPU spike. The agent detects the anomaly, diagnoses the root cause as legitimate traffic (not a loop), and applies a vertical scaling action within safety limits.

Scenario 2: High-Risk Intervention (Human-in-the-Loop)

Here, an agent proposes a destructive action (deleting a dataset) or a high-cost change. The Risk Engine flags this, pausing execution until a human admin approves via Slack or the Dashboard.

Scenario 3: Change Set Lifecycle

Every action in SmartSRE passes through a strict state machine to ensure auditability and consistent rollback capabilities.