The Cast AI blog
Guides, tutorials, and tips on Kubernetes automation, from cost optimization to cloud security and everything in between.
Idle Kubernetes Resources: How to Find and Eliminate Them at Scale
Kubernetes idle capacity is the gap between what workloads request and what they actually use. A rightsize, consolidate, delete sequence reclaims it.
How to Increase GPU Utilization in Kubernetes (5 Proven Fixes)
Put idle GPU capacity to work. This guide covers the metrics, sharing methods, scheduling changes,…
CastPilot MCP: put your agent to work on production problems
CastPilot MCP is now generally available, giving your AI agent read-only access to Cast AI…
Karpenter Enterprise Meets Azure – Karpenter, Your Way: Self-Hosted or NAP-enabled AKS
Karpenter gives AKS teams flexible node provisioning and consolidation. Whether you operate Karpenter yourself or…
Database Caching, Connection Pooling, and AI Index Recommendations in One Tool
The Problem Nobody Wants to Own Someone on your team owns the database. It probably…
New Cast AI Console Experience
We rebuilt the Cast AI console around how you work, not how we ship features.
AKS Cost Optimization: The Real Fixes Under the Hood – NAP, VPA, Spot, and Commitments
AKS clusters often carry the worst utilization metrics in cloud compute ā but there’s no…
Cast AI vs SkyPilot: Two Approaches to Finding GPUs Across Clouds
SkyPilot and Cast AI take different approaches to GPU infrastructure, and the right fit depends…
Kubernetes Configuration Drift: How to Detect It and What It Costs You
Kubernetes configuration drift occurs when actual cluster state diverges from what Git or Helm declares,…
GPU Job Queueing with Kueue and DRA: Scheduling AI Workloads Without Idle Capacity
Kueue and DRA help Kubernetes manage GPU capacity more efficiently by queuing workloads until resources…
4.8/5 250+ reviews