Data Delivery

Once we have the point of sale, receipt data and the mapper files in place, we are ready to use this data as input to our analytic models, so we have meaningful dashboards to get insight from. When it comes to PRIME modules, we can divide them into two big categories, as explained below.

Platform Modules

Modules which are generated end-to-end via VDH are a part of this, this also includes solutions.

For example, the module called Basic Overviews is not a solution but is a part of this category because it is a standard module (created in every client’s environment) and it is generated completely on VDH (does not require a lot of processing power), Product 360 has both these attributes as well but it is a solution (because it is used on direct client dashboards for insight, while basic overviews serves more to PRIME team for internal checks).

When creating a new platform module, we must make sure that we are in Standardized Output, and that the name of the project must be the same as every S3 export. For example, above-mentioned modules are mapped on S3 as /standardized_output/basic_overviews/week_total_data/ or /standardized_output/product_360/total_data/ and so on.

Machine Learning Solutions

Solutions that are created by machine learning libraries and require much more processing power than the usual ones, we run such models on separate, powerful clusters for faster processing time.

These solutions are written in Scala and are not a part of the platform by any mean. In order to run a machine learning solution, you must point to the appropriate JAR file on S3 through EMR. At this moment, we have three such solutions:

1. Promotion Effectiveness

2. Shopper Trips

3. Store Segmentation

These solutions are scheduled to run automatically, and normally only the output of these results is processed in VDH. When creating a solution in VDH which contains any of the above mentioned, we must make sure that we are in Standardized Output project, ML Solutions subproject, and as for S3, the export is not limited to standardized output.

Deliverables

This part of the platform usually contains mostly exports, because the results are directly presented to the clients. It is made of:

• Algemeen

• Promotie

• Formule

• Categorie Management

The names stay in client favorable language (for example, this is in Dutch, as is the client). These projects are maintained by data delivery team together with the client team.