Self-Host Confident AI

Confident AI can be self-hosted in your organization’s infrastructure through a containerized deployment. This deployment model ensures that no data ever leaves your data center, providing enhanced security and compliance for organizations with strict data governance requirements.

Our self-hosted solution includes enterprise Single Sign-On (SSO) for secure and seamless authentication.

Deployment Process

The self-hosting deployment process consists of the following steps:

1. Acquire Enterprise License: Contact our sales  team to obtain an enterprise software license for self-hosting.

2. Receive Deployment Package: You’ll receive a comprehensive deployment package including:

   • Docker images
   • Kubernetes manifests
   • Configuration templates
   • Deployment documentation

3. Configure Infrastructure: Set up the required PostgreSQL and Redis instances according to the specifications.

4. Integrate Identity Provider: Configure your SAML or OAuth 2.0 identity provider:

   • For SAML: Configure the service provider settings in your IDP
   • For OAuth 2.0: Set up the appropriate OAuth application and endpoints

5. Deploy Application: Deploy the Confident AI containers using the provided Kubernetes manifests or Docker Compose files.

6. Verify Deployment: Run through the verification checklist to ensure all components are functioning correctly.

System Requirements

Before deploying Confident AI, ensure your infrastructure meets the following requirements:

Database

• PostgreSQL 14+: Required for storing application data, user information, and evaluation results
• Minimum specifications:
  • 4 CPU cores
  • 8GB RAM
  • 100GB storage (adjust based on expected data volume)

Caching

• Redis 6+: Required for session management and caching
• Minimum specifications:
  • 2 CPU cores
  • 4GB RAM
  • 20GB storage

Identity Provider (IDP)

• SAML 2.0 or OAuth 2.0 compatible identity provider for Single Sign-On (SSO) integration
• Supported providers include:
  • Azure AD
  • Okta
  • OneLogin
  • Ping Identity
  • Custom SAML/OAuth implementations

Infrastructure

• Kubernetes cluster or container orchestration platform
• Load balancer for traffic distribution
• Persistent storage for database and file storage
• Network access to required external services (if applicable)

Support and Maintenance

Our enterprise support team provides assistance throughout the deployment process and ongoing maintenance:

• Initial deployment support
• Regular security updates
• Performance optimization guidance
• Troubleshooting assistance

Timeline

The typical deployment timeline is 1-2 weeks, depending on your organization’s infrastructure complexity and security requirements. This includes:

• Infrastructure setup: 2-3 days
• Application deployment: 1-2 days
• IDP integration: 1-2 days
• Testing and verification: 2-3 days

Contact our team at support@confident-ai.com to discuss your specific self-hosting requirements.