AI Recommendations
Introduction
As you scale with Kubernetes, you may come across the dilemma of inefficiency:
- Over-provisioned clusters
- Workloads running at, let's say, 20% utilization
- Soaring cloud bills
Devtron’s AI-powered initiative redefines how Site Reliability Engineers (SREs) and DevOps teams interact with their infrastructure. It monitors, reasons, and acts on cost inefficiencies using explainable AI.
It operates across two modules to implement rightsizing actions:
- Notifications - Automated optimization insights of your Kubernetes resources.
- Runbooks - Predefined or AI-generated remediation workflows.
User Personas
| Persona | Role |
|---|---|
| SRE / DevOps Engineer | Primary user for AI recommendations and runbook automation. |
| Superadmin | Reviews approvals and monitors audit trails. |
| Developer | Queries workloads and costs through AI Chat. |
Associated Modules
Notifications
Sends intimation regarding potential optimization across clusters to save costs.
- Users can Approve, Reject, or Revert recommendations.
- Each action links to a relevant runbook for remediation.
Click here to know more about Notifications.
Runbooks
Defines YAML-based remediation actions.
- Supports indefinite and time-bound approvals.
- Includes per-cluster execution and audit tracking.
Click here to know more about Runbooks.
Audit Logs
Maintains a full record of all user and AI-driven actions.
- Provides audit trail of runbook.
- Filterable by user, module, and action type.
Additional Resources:
- Watch Devtron's AI Capabilities.
- Ask Devtron Expert - A simple chat interface for queries and analytics (accessible from the top-right of your screen).
- Devtron Intelligence - An AI agent that helps you will troubleshooting of workloads.