Create a project
From Home > Projects, click Create Project.
Once the project is created, you can see the following tabs:
Details
The Details tab gives information about the project resources and resource usage, including:
- which resource pool the project belongs to,
- what resource quotas it has, and
- if it is allowed to use GPU.
Read more: Resource management
Data
If you have not already done so, Add Data to the catalog.
- If you Add Data from within the project, the data is owned by the project and no further action is necessary.
- Otherwise, once your data is available in the catalog, Link to Project.
In the example displayed in the screenshot, the data has been added from within the PredictiveMaintenance project, and the project is the owner of the data.
From within the Data tab, under Actions, you can start Auto ML or, by pressing the three-dots menu, submit the data to Feature Mart. Feature Mart attempts to optimize your data by creating new columns better adapted to your use case.
Connections
Alternatively, your data source may be external, in a database or other cloud storage. The Connections tab shows all the data connections available to the project. Click Add Connection to add a new one.
Read more: Connect to your data
Content
Once your data is ready, you can take action within the Content tab:
- Start Auto ML: Start Auto ML based on the data set.
- Start Auto FE: Start an Auto Feature Engineering run to generate an optimum set of features based on your original data.
- Create Workflow: Develop a workflow to visually transform, model, or score your data in a completely custom way.
- Create App: Create an app to present your data, models and other results as a dashboard.
- Open Notebooks: Open our platform integrated notebooks in a new tab to collaborate on the project using Python code.
- Import from ZIP
Access
The Access tab makes it possible for you to invite other colleagues to work with you on the project. Click Add User and select a User and Permission:
- Viewer - can read the project files, but not change them in any way
- Contributor - can make changes, read and write to the project
- Owner - has the right to delete the project
Deployments
With deployments, one can enable content created within the project to be integrated into existing business processes. This can happen in multiple ways, such as creating an automated schedule when a workflow should be executed, or exposing a workflow as a REST endpoint so that other applications can leverage it.
Read more: Deploy a project
Workspaces
Looking for a lightweight, but powerful source code editor to help you contribute Python code to your project?
A workspace provides exactly that: a source code editor, connected to a specific project. It runs in your browser and it comes with all the features you expect:
- Code completion
- Debuggers
- Version control with Git
- A large set of available extensions
Read more: Workspaces
Coding environments
Coding environments are self-contained language environments with packages and versions. In Altair AI Cloud, you can create code environments for Python. You can use them in notebooks, workspaces, and Execute Python operators in the workflow designer.
Read more: Coding environments
Planning
The Planning tab implements an integration with Atlassian Jira and Confluence, de facto standards in project management.