Agile AI Platform Architecture with the Agile Cloud Manager
Part 8 of 10: Azure Example Appliance
The entire transcript of this video is given below the video so that you can read and consume it at your own pace. We recommend that you both read and watch to make it easier to more completely grasp the material.
The AgileCloudInstitute.io web site contains working examples of software-defined A.I. platform appliances that you can deploy into multiple cloud providers. You can seed your A.I. platform development projects by starting with these working examples.
One of the working examples enables you to create the seed of an Azure A.I. platform appliance.
The architecture section of the AgileCloudInstitute.io web site includes videos and text describing this appliance, and the marketplace section of the AgileCloudInstitute.io web site includes hands-on training for using our free templates to deploy this appliance into Azure.
This example appliance has three systems.
A core system in this example includes a foundation that includes items that are shared by other services and by other systems. The foundation in this example includes a shared storage account and a shared Purview account.
The core system also includes a number of services, including a service that manages a Synapse workspace, and other services that manage other things.
There is also a data engineer system in this example. The data engineer system does not have its own foundation, and instead uses some of the shared resources from the core system’s foundation.
The data engineer system in this example contains a service that manages a Synapse workspace for data engineers. There are also other services defined in the data engineer system in this example.
A data scientist system is also included in this example. The data scientist system in this example also shares some of the resources from the foundation of the core system. And the data scientist system also includes:
- A service for creating a machine learning workspace.
- And other services.
You can easily add a service for creating another synapse workspace for data scientists. This example works within Azure’s limitation of only allowing two synapse workspaces per account. So this example omits a data scientist synapse workspace in order to make it easier for you to get this working example up and running quickly without having to use multiple Azure accounts.
Preprocessors and postprocessors are also included in each of the systems in this example appliances. Preprocessors and postprocessors do things that templates cannot do. Some examples of what preprocessors and postprocessors do include:
- Network administrative tasks.
- Copying data.
- Running automation.
- Deleting data.
- And removing protected resources.
The way that you would implement the A.I. project management items described in this video into this working example would be to build out the software definition of the systems defined in this working example.
For example, a best practice is to create a machine learning workspace for each of your A.I. projects. That way, you can create software definitions of everything that you want to be included in each of your projects. Even role-based access control to manage all the project components. Then, each time you create a new project, all you have to do is to run the Agile Cloud Manager C.L.I. commands that create a new machine learning workspace fully provisioned with everything needed, and with role-based access control already set up.
This working example is meant as a starter. It is meant to seed your project. So that you can get a working example up and running quickly on day one. And then you can iterate the working example, to evolve it into something that fits the unique needs of your evolving organization.
Proceed to Part Nine: AWS Example Appliance
Back to Series Table Of Contents: Agile AI Platform Architecture With Agile Cloud Manager