Raw datasets act as containers for data that you import from various source systems. You can also use these raw datasets for reference when you create engineered datasets.

You can use engineered datasets to modify the downloaded data based on your requirements and upload it back. You can add to the existing data, remove from it, or perform a data cleansing on the whole data.

It is a building block that helps in determining the needs attributes that are used by the runtime application in recommending the best matching products.

Solution is a building block that helps in determining the product attributes that are used by the runtime application in recommending the best matching products. You can assign a configurable product to a solution to further define a product category. Solutions and configurable products are always a part of a product category.

You can define the source system to extract data from, the time range of the data, and the pricing procedures that are applicable to it.

You can then upload this template back into the system and use its data for machine learning trainings.

You can choose either raw or engineered datasets, or both as reference datasets. You can download this document, cleanse the data in it or make changes, and upload it back to the system.