Skip to content
CloudPivot

Scheduler Engine · CloudPivot platform

Your cloud doesn't work weekends.
Stop paying like it does.

The Scheduler Engine learns each resource's usage pattern, proposes an off-hours schedule priced in dollars, and applies it in one click — across ASGs, ECS services, EC2, and EKS node groups.

How the engine decides

From usage pattern to applied schedule in three steps.

1 · Observe

Weeks of utilization, request, and deploy-time signals per resource — with metric-confidence badging, so a proposal built on thin data says so.

2 · Propose

A concrete schedule — say, Fri 19:00 → Mon 06:00 off — with the projected dollar savings computed from your actual rates, not list price.

3 · Apply & watch

One click replaces hand-written cron. Every scheduler event logs real dollars saved, and savings-leakage detection catches a schedule that quietly stops firing.

Coverage

One engine, every schedulable dollar.

Auto Scaling Groups

Scale-to-zero or scale-to-floor windows, with every scheduler event priced — not a $0 placeholder.

ECS services

Desired-count schedules for services that nobody calls at 2am, applied without touching task definitions.

EKS node groups

Off-hours node-group scheduling extends the same engine to Kubernetes — the fee-trap-aware way.

EC2 instances

Plain instances get stop/start windows, burstable-credit economics included in the math.

RDS dev & test

Stops that survive AWS's 7-day auto-restart — the engine re-stops what AWS wakes up, a correctness detail most schedulers miss.

Commitment aware

Savings Plans and RI coverage are modeled against proposed schedules, so a stop never strands a commitment you already paid for.

Weekend math

Fri 19:00 → Mon 06:00 is 59 hours. Every week.

35%

of the week is weekend — hours most non-production fleets bill at full price for nobody

59 hrs

of sleep per resource per week from the default weekend window alone

+ nights

usage-pattern proposals extend the same engine to weeknights and quiet hours

Trusted automation

Schedules your platform team will actually sign off on.

Dry run first

Every schedule starts as a preview: which resources, which windows, how many dollars.

Approval in Slack

Proposals route to the resource owner. Approve, snooze, or dismiss without leaving your workflow.

One-click undo

Any schedule can be reverted instantly. Weekend experiments shouldn't need a change-review meeting.

Give your cloud the weekend off.

Connect an account and the engine proposes its first schedules from your real usage — free to see, one click to apply.