Savings that apply themselves
The recommendations act on their own. Schedules, spot conversion and autoscaling run continuously instead of waiting for a ticket.
Everyone knows dev boxes should sleep at night and batch jobs should run on spot. The change never gets made because it is nobody's sprint.
CloudSitter
Schedule instance start/stop on a weekly grid with a projected savings figure before you commit.
Non-production infrastructure rarely needs to run overnight or at weekends, but the schedule never gets implemented because it is nobody's priority. CloudSitter sets start/stop windows on a weekly grid and shows the projected saving before you commit to it.
- Weekly grid for start/stop windows
- Projected saving shown before you apply the schedule
- AWS, GCP and Azure
Move Auto Scaling Groups onto spot capacity automatically, with savings history and one-click CloudFormation deploy.
Spot capacity is the biggest discount in cloud compute and the one teams avoid because interruptions look risky. AutoSpotting moves Auto Scaling Groups onto spot automatically and keeps a savings history, and it deploys through a one-click CloudFormation stack.
- Converts Auto Scaling Groups to spot capacity
- Savings history per group over time
- One-click CloudFormation deployment
ASG configuration
Per-group spot-to-on-demand ratio and instance-type diversification.
Blanket spot conversion is the wrong answer for a group serving live traffic. Each Auto Scaling Group gets its own spot-to-on-demand ratio and instance-type diversification, so the risk profile is set per workload rather than globally.
- Per-group spot to on-demand ratio
- Instance-type diversification to reduce interruption risk
- Currently AWS
Kubernetes autoscaling controller
HPA and VPA management, node optimization, rebalancing and pod rightsizing in-cluster.
Cluster efficiency depends on decisions at several layers at once: pod requests, replica counts and the node pool underneath. The controller manages HPA and VPA, optimizes and rebalances nodes, and rightsizes pods from inside the cluster.
- HPA and VPA management
- Node optimization and rebalancing
- Pod rightsizing based on real usage
Cost automation, answered.
FinOps finds and quantifies the savings. Cost automation applies them: CloudSitter runs start/stop schedules, AutoSpotting converts Auto Scaling Groups to spot capacity, and the Kubernetes controller rightsizes continuously, all without a ticket.
Risk is configured per workload rather than globally. Each Auto Scaling Group has its own spot-to-on-demand ratio and instance-type diversification, so latency-sensitive groups keep a guaranteed floor of on-demand capacity.
Yes. CloudSitter shows a projected savings figure for a schedule before you commit to it, and AutoSpotting keeps a savings history per group after conversion.
Through a one-click CloudFormation deployment, with savings history reported back into XamOps.
Cost automation is one module of 8.
Ready to automate cost automation?
30-minute walkthrough. We connect to a sandbox and show this module running against real infrastructure.