Integrator Documentation

Documentation for engineers integrating AI Cluster Runtime (AICR) into CI/CD pipelines, GitOps workflows, or larger platforms.

Audience

This section is for integrators who:

• Build automation pipelines using the AICR API
• Deploy and operate the AICR API server in Kubernetes
• Create custom recipes for their environments
• Integrate AICR into GitOps workflows (ArgoCD, Flux)

Documents

Document	Description
Automation	CI/CD integration patterns for GitHub Actions, GitLab CI, Jenkins, and Terraform
Data Flow	Understanding snapshots, recipes, validation, and bundles data transformations
Kubernetes Deployment	Self-hosted API server deployment with Kubernetes manifests
EKS Dynamo Networking	Security group prerequisites for Dynamo overlays on EKS
AKS GPU Setup	AKS prerequisites: Kubernetes 1.34+ (DRA GA), GPU driver setup, DRA configuration
Recipe Development	Creating and modifying recipe metadata for custom environments
Validator Extension	Adding custom validators and overriding embedded ones via --data

Quick Start

API Server Deployment

# Deploy API server to Kubernetes
kubectl apply -k https://github.com/NVIDIA/aicr/deploy/aicrd

# Generate recipe via API
curl "http://aicrd.aicr.svc/v1/recipe?service=eks&accelerator=h100"

CI/CD Integration

# GitHub Actions example
- name: Generate recipe
  run: |
    curl -s "http://aicrd.aicr.svc/v1/recipe?service=eks&accelerator=h100" \
      -o recipe.json

- name: Generate bundles
  run: |
    curl -X POST "http://aicrd.aicr.svc/v1/bundle?bundlers=gpu-operator" \
      -H "Content-Type: application/json" \
      -d @recipe.json \
      -o bundles.zip

Related Documentation

• Users: See User Documentation for CLI usage and installation
• Contributors: See Contributor Documentation for architecture and development guides