Kubernetes and data
Kubernetes OptimizationRight-size workloads with Karpenter or Cast AI’s autoscaler.
Cast RadarCost, error rate and latency across every cluster.
Storage AutoscalerGrow volumes before they fill. Tune IOPS to demand.
Database OptimizerFind slow queries. Cache reads and pool connections.
GPU and AI
OMNI ComputeUse GPUs in other clouds and regions when yours run out.
GPU sharingShare one GPU across workloads, and rightsize them.
Agents and operations
Agentic workflowsAgents open PRs for infra fixes. You review and merge.
OpsPilotAsk about pods, costs and health in plain English.
MCP serverQuery cluster cost and health from your AI assistant.
Environments
Cast AI AnywhereOptimize on-prem and bare metal with your autoscaler.
CloudsRuns on AWS, GCP, Azure and Oracle Cloud.
On-premisesSupports OpenShift, Tanzu, kOps and bare metal.
Cost and performance
GPU and AI
- Access and optimize GPU capacitySource GPUs across clouds and improve utilization.
Operations
- Reduce manual operationsAutomate routine tuning within your policies.
- Verify production impactStart read-only and validate changes before expanding.
Customers
Pricing
Explore
Learn
Community
Cast AI
About usOur team, mission, and story.
NewsroomThe latest from Cast AI.
EventsFind us online and in person.
Contact
CareersBuild what comes next with us.
Partner programGrow with Cast AI.
Contact usFind the right team to talk to.

October 29th, 1PM ET

Live panel

Workloads and Nodes, Tuned Together: Kubernetes Automation in Practice

2026/10/29 16:00 GMT

–

days

–

hours

–

minutes

–

seconds

Most Kubernetes clusters pay for far more capacity than they use. Bret Fisher and Steve Fraser will show how Cast AI sizes workloads from real usage, scales nodes to match, and how teams roll this out on production clusters one step at a time.

Why Attend?

Average CPU utilization across Kubernetes clusters is around 8%, and 69% of clusters request more CPU than their workloads use. If you run Kubernetes, some of this probably sounds familiar:

Resource requests still carry the headroom someone added at launch
The node autoscaler scales to requests, so padded requests mean extra nodes
Half-empty nodes stay up because the pods on them are stateful or long-running
Spot stays off after one interruption took something down
Automation stays off production until someone proves it’s safe

We’ll walk through how to fix each of these and what the rollout looks like, starting in read-only mode and enabling automation cluster by cluster.

Source: Cast AI 2026 State of Kubernetes Optimization report

🎁 BONUS: Everyone who registers is entered into the raffle for AirPods Pro 3 and exclusive Cast AI swag.

What You’ll Learn

How to set CPU and memory requests from real usage and resize pods in place, without restarts
How the node autoscaler picks cheaper instances, packs pods tighter and handles Spot
How live migration moves stateful and long-running pods with minimal disruption, so underused nodes can be consolidated
How to add these optimizations on top of Karpenter or Cluster Autoscaler without changing how nodes get provisioned
What savings and utilization gains teams see once automation is running

Live Discussion Format

Bret and Steve will talk through how teams actually run this in production: what they rightsized first, where automation needed guardrails, and what the savings looked like after a few months. Bring your own cluster questions. We’ll leave plenty of time for Q&A.