Target tenant resources
Users must have the COLEMANAI-User or the COLEMANAI-Administrator role to use and edit resources in the target tenant.
This table describes the resources that can be included in the global packages.
| Resource | Category | Description | Live instance | Reference instance | Overwrite |
|---|---|---|---|---|---|
| Dataset | Data collection | A dataset is the data that is used as input in a quest to build a model. The import in the target tenant includes the definition of the dataset from Infor Data Lake. | The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Support is not available at this time |
| Groups | Data collection | The groups bind the multiple datasets together with a unique name. This is a mandatory step that is required for the successful import of the Optimization quest into the target tenant. | The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Support is not available at this time |
| Quest | Machine learning | A quest is the flow of activities that build the machine learning model.
The import in the target tenant includes the training and production quest definition. |
The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Support is not available at this time |
| Endpoint | Machine learning | An endpoint is the deployed machine learning model (REST API) that is invoked to get real-time predictions.
The import in the target tenant includes the endpoint definition. |
The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Support is not available at this time |
| Custom Algorithm | Machine learning | A custom algorithm is a user-defined source code that is used as an algorithm to train a machine-learning model.
The import in the target tenant includes the custom algorithm docker image and the source code files package. |
The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Supported. Users can overwrite changes in the live instance with content from the reference instance. |
| Quest | Optimization | A quest is the flow of activities that build the optimization model.
The import in the target tenant includes the design and production quest definition. |
The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Support is not available at this time |
| Endpoint | Optimization | An endpoint is the deployed optimization model (REST API) that is invoked to get real-time solutions.
The import in the target tenant includes the endpoint definition. |
The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Support is not available at this time |
| Custom algorithm | Optimization | A custom algorithm is a user-defined source code that is used to solve optimization problems.
The import in the target tenant includes the custom algorithm docker image and the source code files package. |
The content is fully editable to meet business requirements. | The reference instance is a read-only instance that supports review of the latest package content. | Supported. Users can overwrite changes in the live instance with content from the reference instance. |