Match IQ Workflow

The workflow is a sequence of steps for matching entities by using machine learning techniques.

The high-level workflow of how to use a model for matching is as follows:
  1. The first step is to create a model flow by selecting the entity type and attributes.
  2. After the preparing data stage is complete, you need to train the model by answering questions.
  3. The trained model then matches records within the tenant. You can download and review the match results. If you are satisfied with the results, you can approve the model.
  4. Once the model is approved, you need to publish the model where you need to identify the scope for the model. The scope can be Internal, External, or both.

When matching is triggered in the tenant, a published model with an internal scope also participates in matching along with the match rules that are defined in the tenant. As a result, in the Potential Matches screen in Hub, a data steward can view both rule based and Match IQ model based recommendations.