Deeploy Installation & integration options
There are multiple ways to integrate Deeploy in your environment. In short:
- Use the Deeploy managed (multi-tenant) SaaS environment
- Use the Deeploy managed (single tenant) Enterprise SaaS environment, which is available via cloud marketplaces and via our website. We guide you through the structured integration steps for your specific requirements.
- Use Deeploy via a self managed installation, this is where the remainder of this section will focus on.
Prerequisites for self managed installations
- The Kubernetes CLI (kubectl)
- Helm v3
- Cloud or bare metal resources
- An SMTP service and credentials (e.g, Sendgrid), check here the required SMTP config for the Deeploy installation
- An DNS management service (e.g., see), check the required DNS config and the required TLS config for the Deeploy installation.
After the prerequisites are in place, the Deeploy Helm Chart can be configured and deployed, as described here.
Supported Cloud platforms
- Azure, also check out our offerings in the Azure marketplace
- AWS, also check out our offerings in the AWS marketplace
- GCP
- OVHcloud (EU)
- Scaleway (EU)
Bare metal installations
Deeploy can run on your own servers, we describe a starting point for the installation here. For more details contact us.
SSO with OIDC
Deeploy supports the following SSO providers by default:
- Microsoft
- Okta
Check our SSO page for more details. Don't see your SSO provider? Contact us for more details.
Integrations
Deeploy is an open platform that aims to integrate with many tools and platforms in the ML ecosystem. Have a look at the following integrations:
Model training and version control
- Databricks
- MLflow
- Huggingface
Messaging
- Slack
- Microsoft Teams