Catch Up on the Azure Communication Services Fundamentals Series
May 6, 2025Microsoft Surface Platinum Partners: Empowering success through endpoint strategy
May 6, 2025
We are pleased to announce that hubs and workspaces is now generally available on Azure machine learning allowing users to use hub for team’s collaboration environment for machine learning applications.
Azure Hubs and Workspaces provide a centralized platform capability for Azure Machine Learning. This feature enables developers to innovate faster by creating project workspaces and accessing shared company resources without needing repeated assistance from IT administrators.
Quick Model Building and Experimentation without IT bottleneck
Hubs and Workspaces in Azure Machine Learning provide a centralized solution for managing machine learning resources. Hubs act as a central resource management construct that oversees security, connectivity, computing resources, and team quotas. Once created, they allow developers to create individual workspaces to manage their tasks while adhering to IT setup guidelines.
Key Benefits
- Centralized Management: Hubs allow for centralized settings such as connectivity, compute resources, and security, making it easier for IT admins to manage resources and monitor costs.
- Cost Efficiency: Utilizing a hub workspace for sharing and reusing configurations enhances cost efficiency when deploying Azure Machine Learning on a large scale. There is a cost associated with setting separate firewall per workspace which scales up as the number of workspaces go up. With hubs, only one firewall is needed which extends across workspaces saving cost.
- Resource Management: Hubs provide a single pool of compute across workspaces on a user level, eliminating repetitive compute setup and duplicate management steps. This ensures higher utilization of available capacity and fair share of compute resources.
- Improved Security and Compliance: Hubs act as security boundaries, ensuring that different teams can work in isolated environments without compromising security.
- Simplified Workspace Creation: Hubs allow for the creation of “light-weight” workspaces in a single step by an ML professional.
- Enhanced Collaboration: Hubs enable better collaboration among data scientists by providing a centralized platform for managing projects and resource
How to get started with Hubs and Projects
There are different ways to create hubs for users. You can create hubs via Azure portal, with Azure Resource Manager templates, or via Azure Machine Learning SDK/CLI. Hub properties like networking, monitoring, encryption, identity can be customized while creating a hub and can be set depending on org’s requirements. Workspaces associated with a hub will share hub’s security, connectivity and compute resources.
While creating hubs via ML Studio is not supported currently, once hub is created users can create workspaces which get shared access to the company resources made available by the administrator including compute, security and connections.
Besides ML Studio, workspaces can be created via Using Azure SDK, Using automation templates, Using Azure CLI.
Secure access for Azure resources
For accessing data sources outside hubs, connections can help make data available to Azure machine learning. External sources like Snowflake DB, Amazon S3 and Azure SQL DB can be connected to AML resources.
Users can also set access permissions to the azure resources with Role based access controls. Besides default built-in roles, users can also create custom roles for more granular access.
To conclude, the General Availability of Azure Machine Learning Hubs and Workspaces marks a significant milestone in our commitment to providing scalable, secure, and efficient machine learning solutions. We look forward to seeing how our customers leverage this new feature to drive innovation and achieve their business goals.
For more information on hubs and workspaces in Azure machine learning, please refer the following links.