Time to complete: 5 minutes
What you’ll get: xpander.ai running on your own Kubernetes cluster with full data control
What you’ll get: xpander.ai running on your own Kubernetes cluster with full data control
Quick K8s Deployment
1
Create Environment
Create a self-hosted environment in the xpander.ai Console:
- Go to app.xpander.ai/environments
- Click “Add Environment”
- Enter your server FQDN (e.g.,
xpander.my-company.com
)
The platform provides your complete Helm installation commands!
2
Install with Helm
Run the provided Helm command from the UI. This is just an example. Make sure you provide your OpenAI/Anthropic keys.
3
Verify
Check deployment:All pods should show
Running
status.What You Deployed
xpander self-deployed engine
Agent controller, AI gateway, MCP server, and worker nodes
Your Data Stays Private
All conversations and business logic remain in your cluster
Outbound Only
No inbound connections - only connects out to the xpander.ai cloud
Ready to Use
Easily deploy all your existing or new agents and MCP servers to your self-hosted environment
Prerequisites
Required before you start
Required before you start
- Kubernetes 1.20+ with kubectl access
- Helm 3.12+ installed
- Ingress Controller (NGINX recommended)
- Storage Class for persistent volumes
- Outbound HTTPS access to
166.117.85.46
and15.197.85.80
Next Steps
Your self-hosted xpander.ai is ready! Now build:- Visual Agent Builder to Code - Create agents on your infrastructure
- OpenAPI Spec to MCP Tool - Deploy MCP servers to your cluster
- Connect to Slack - Team integrations
Troubleshooting
Pods not starting
Pods not starting
Check storage and permissions:
kubectl -n xpander describe pod <pod-name>
Can't access endpoints
Can't access endpoints
Verify ingress and DNS:
kubectl -n xpander get ingress
Need help
Need help
Full deployment guide: Self-Hosted Deployment