Hello Team,

Since OpenShift AI requires a lot of resources (8 CPUs, 32Gi of memory) and, on top of that, other operators for its full capabilities (Service Mesh, Serverless, Pipelines), I think that it would be beneficial for customers to be able to schedule OpenShift AI's infrastructure workload on the "infra" MachineSet, as it is already the case with other embedded products, even with Service Mesh, Serverless, and Pipelines themselves. In this way, if they choose to take advantage of this capability, customers would have a lot of free worker resources to spend on their workloads, avoiding unnecessary extra licenses.