An Enrichment takes data in and outputs a predicted property about that data. Sometimes called a classifier, examples of enrichments include predicting sentiment on text, classifying text by topic, or labelling an object in a photo.
To generate predictions about your data you first need an Enrichment Model, Graft provides a number of models you may select from or you may choose to create your own custom models from labeled data, you can also create initial Enrichment models using the Graft Bootstrapping functionality which takes unlabelled data and sample label names you provide to generate those initial labels which can be reviewed and corrected using Active Learning.
Once created, enrichment models can be used on any entity you have created in your project.