Overview

The SRE Agent helps streamline incident response by analyzing configured Incident Workflows and recommending the most relevant one based on the real-time context of each incident. Recommendations include an explanation of the reasoning and a potential ranked list of workflows.

Use the SRE Agent in:

• Operations Console (AIOps customers only): On the SRE Agent tab within an incident view, you’ll see the Agent’s recommendation summary and reasoning, along with a “Run recommended workflow” button.
• Incident Detail's Page: On the SRE Agent tab in the side panel, you’ll see the Agent’s recommendation summary and reasoning, along with a “Run recommended workflow” button.
• Slack: Recommended workflow and run options will appear within your associated incident channel.

How it works

1. Initial Contextual Analysis
   The SRE Agent automatically reviews the incident context, including event payloads, historical incident patterns, service integrations, and conversation history.

2. Accessing Recommendations
   Users are presented with a "Recommend a Workflow" nudge. Selecting this prompt triggers the agent to evaluate all configured Incident Workflows.

   

3. Top & Alternative Recommendations
   The agent identifies the workflow that best matches the current situation and presents it as the top recommendation, along with other relevant alternatives. Each recommendation includes a brief reasoning for why it was selected.

   ⚠️

   Recommendations only appear if:

   • Incident Workflows are properly configured in your account.
   • The SRE Agent identifies a high-confidence match between the incident and available workflows.

4. Execution
   Review the agent's logic and select the "Run recommended workflow" nudge to execute the automation.

   

5. Refining via Chat
   If the top recommendation isn't what you need, you can use natural language to ask the agent for a specific alternative (e.g., "Recommend the rollback workflow instead"). The agent will then surface that specific nudge for execution.

Important Notes

📘

Execution Guidelines

• Execution Limit: Each recommended workflow can only be executed once per incident.
• Re-run Notifications: If a user attempts to trigger a workflow that has already been run, the SRE Agent will send a message in the Operations Console, Incident Detail's page, and Slack indicating that it cannot be re-executed.
• Success Confirmation: Once a workflow is triggered, the SRE Agent will display a confirmation message to verify successful execution.
• Timeline Audit: All workflow executions are automatically logged in the Timeline tab of the Incident Details page for auditing and visibility.