HKE: Hanzo Kubernetes Engine

Kubernetes won. The orchestration debate is over. But "use Kubernetes" and "operate Kubernetes" are very different activities.

Operating Kubernetes in production means: control plane upgrades, etcd maintenance, node auto-scaling, CNI networking, ingress configuration, certificate rotation, RBAC policy management, resource quota enforcement, PV provisioning, and monitoring all of it. That is a team, not a tool.

Today we are launching HKE: Hanzo Kubernetes Engine. Managed Kubernetes that eliminates the operational burden without limiting what you can do.

What HKE Manages

Control Plane

• Fully managed, highly available control plane
• Automatic version upgrades (configurable maintenance windows)
• etcd backup and recovery
• API server scaling based on cluster size

Node Management

• Auto-scaling node pools (scale to zero supported)
• Mixed instance types per pool (CPU, GPU, high-memory)
• Automatic OS patching and security updates
• Spot/preemptible instance support for cost optimization

Networking

• CNI networking with network policies
• Automatic ingress with TLS termination
• Service mesh integration (optional)
• Private clusters with no public endpoints

Storage

• Automatic persistent volume provisioning
• Storage class management
• Volume snapshots and backup
• CSI driver support

Security

• RBAC preconfigured with sensible defaults
• Pod security policies enforced
• Image vulnerability scanning
• Secrets encryption at rest
• Audit logging

GPU Support

First-class GPU node pools for AI/ML workloads:

• NVIDIA A10G, A100, and H100 instances
• Automatic GPU driver installation
• GPU sharing and time-slicing
• CUDA toolkit pre-installed

Pricing

• Control plane: Free (we do not charge for management)
• Nodes: Pay for compute at standard instance rates
• Support: Included in Hanzo Cloud subscription

No markup on compute. No hidden management fees.

---

Zach Kelling is the founder of Hanzo AI (Techstars '17).