Self hosting Patronus AI guide
Complete guide for deploying Patronus AI in your own infrastructure
This guide walks you through the complete process of self-hosting Patronus AI in your own infrastructure. Follow these steps in order to ensure a successful deployment.
Deployment overview
Understand the architecture
Review the Architecture Overview to understand how Patronus AI components work together and plan your deployment.
Prepare your environment
- Review the Prerequisites to ensure your infrastructure meets all requirements for hosting Patronus AI.
- Test your identity provider (IdP) authentication using the Vouch-proxy configuration test guide.
- Optionally, deploy the infrastructure base with Terraform.
Configure deployment options
Review the Common deployment options for configuration settings applicable to all cloud providers (Docker Hub, cert-manager, ExternalDNS, pod authentication).
Cloud-specific configurations:
- Deployment in AWS with Amazon Cognito - AWS-specific setup with Cognito authentication
Optional configurations:
- If using external databases (recommended for production), follow the External databases setup guide.
- Configure evaluator authentication for AI model providers using the Evaluators configuration guide.
- Upload and manage datasets using the Datasets configuration guide.
Deploy with Helm
Install Patronus AI using the Helm chart installation guide for your Kubernetes cluster.
Post-installation configuration
- Configure user accounts and authentication flows using the Accounts and authentication guide.
Additional resources
Need help? Contact the Patronus AI team for support with your deployment.