Self-Hosted
Self-Hosted Architecture
In this configuration, both planes are hosted in your organization’s private cloud.
The Self-Hosted configuration provides complete control over the Robust Intelligence cluster’s behavior, including network routing, storage management, and compute resource optimization for model testing jobs.
Note that services may behave differently if alternative configurations are used.
Deployment Process
Your Solutions Architect (SA) will lead the deployment with your MLOps team.
- Phase 1: Scoping
Estimated Duration: 1 session with an RI Solutions Architect and your DevOps Administrator(s)
Your SA will guide you through the system requirements so that you can prepare resources with your internal teams in advance.
By the end of this phase, you should be ready to begin executing the Helm deployment.
- Phase 2: Deployment
Estimated Duration: 3 sessions with an RI Solutions Architect and your DevOps Engineer(s)
At this point, you can begin executing the relevant deployment commands, iterating on configurations as necessary.
- Phase 3: Validation
Estimated Duration: 1 Session with an RI Solutions Architect and your DevOps Engineer(s)
Your SA will guide you through smoke-testing the new Robust Intelligence deployment through a series of validation steps.
This documentation is available for your reference — begin with Requirements below.