A Solution for ML Pipeline in Multi-tenancy Manner

Thanks, but can you please clarify what you mean by the term "tenant"/"tenancy"? In the language of Azure, I understand "tenant" to have a very specific meaning - the uppermost level of the Azure Management Group hierarchy, under which a tenant can have multiple Azure Subscriptions, which can each have multiple Resource Groups etc. I.e. the kind of "tenant" referenced by this doc: https://learn.microsoft.com/en-us/azure/azure-portal/get-subscription-tenant-id

It's not clear to me whether this is the meaning of your term "tenant" or whether you have a more informal meaning along the lines of "different groups of ML practitioners working loosely together within an organisation".

I landed on this page because the specific problem I am trying to solve is:

• Our company has a global federated structure, where we work with different business units that act as their own companies, where each have their own Azure tenancy (in the sense of Microsoft Entra Tenancy I referenced above)
• Our ML project has environments (Azure Resource Groups) split across two of these Azure tenancies (Microsoft Entra tenancies). I.e.
  • Dev environment:   Az Tenant 1 -> Az Subscription 1 -> Az Resource Group 1 -> Az ML workspace 1
  • Prod environment: Az Tenant 2 -> Az Subscription 2 -> Az Resource Group 2 -> Az ML workspace 2
• I want to use an AzureML Registry to share models and environments between Dev & Prod

Is it possible in this scenario using standard Azure AD authentication? Or is the AzML registry not set up for cross-sharing between different Microsoft Entra Tenancies?