Skip to main content

Resource scheduling

PreviewReviewed 21 August 2026
Preview
This guide describes an intended experience that is not generally available. The screenshots use sample data. Workflows, labels and availability can change before release.

This is not Harness AutoStopping. Harness can shut idle resources down. FinomateAI connectors are read-only. This page describes an intended schedule planner: estimate savings for a stop window, then implement the schedule in AWS, Azure, Google Cloud or your orchestrator.

PreviewResource scheduling preview with weekday night and weekend stop windows
Preview. FinomateAI does not execute start or stop. The owner applies the window in the cloud.

Intended workflow​

  1. Scope to env=dev or env=staging. Never production.
  2. Set a window in the organisation's timezone.
  3. Read the estimated saving. Treat it as a model, not a purchase order.
  4. Assign an owner.
  5. Implement Instance Scheduler, Azure Automation, Cloud Scheduler, or a Kubernetes scale-to-zero in your account.

Until this ships, treat idle-shaped Recommendations as advice to implement in the provider.