Overview
The On-Prem AI Infrastructure Platform ships a complete MLOps toolchain ready to use. Kubeflow covers pipelines, notebooks, hyperparameter tuning, and distributed training, KServe serves models in production, and MLflow tracks experiments and manages models. The whole stack is delivered as a one-click AI cluster template.The AI cluster template is a Polystack K8SaaS cluster template.
Deploying it creates a GPU-ready Kubernetes cluster with the full MLOps stack installed.
The AI Cluster Template
One template deploys all of the following:Toolchain
- Kubeflow
- KServe
- MLflow
- Harbor
Model Lifecycle
Develop
Data scientists explore data and build models in Kubeflow Notebooks on GPU nodes.
Train and tune
Kubeflow Pipelines orchestrate training, the Training Operator runs distributed jobs, and
Katib tunes hyperparameters.
Track and register
MLflow records every run and registers the selected model version.
Serve
KServe deploys the registered model as a production inference endpoint.
Next Steps
GPU and HPU Performance Optimization
GPU sharing, multi-node training networking, and accelerator telemetry.
Runtime Vulnerability Scanning
How Harbor and StackRox secure MLOps workloads.
Deploy a Kubernetes Cluster
Deploy clusters from Polystack K8SaaS templates.
Multi-Vendor Accelerator Support
Accelerator choices for AI clusters.