MadKudu pulls data from your system on one end of the pipeline to push scores and segmentation back into your system on the other end of the pipeline.
Several steps happen in between to generate these scores. Because your systems don't necessarily have the same format of data, a step of data preparation and standardization is necessary to format your data according to MadKudu standards to be reused across the platform. This process is called a mapping.
There are 4 different mappings to configure during the implementation of MadKudu or update according to their use cases.