Use machine-learning models

  • 3 minutes

The built-in prediction capabilities of Customer Insights - Data help provide organizations with a starting point to create predictions based on their data. However, as with many predefined options, they might not represent how your organization makes predictions. For example, your organization might want to predict something other than customer churn. Even in scenarios where you want to predict potential customer churn, your organization might structure their subscriptions differently than what the out-of-the-box churn model provides. Alternatively, you might consider different factors when deciding if a customer might churn.

Many organizations might have created machine-learning models that are already making predictions, or some might want to create a model to use in Customer Insights - Data. Customer Insights - Data supports organizations using custom models and managing workflows based on Azure Machine Learning models in the application. Workflows help you identify the data that you want to generate and map the results to your Customer Insights data.

Dynamics 365 Customer Insights – Data supports the following custom models:

  • Azure Machine Learning V2 – You can import your Azure Machine Learning model to make predictions on your customer data.
  • Microsoft Azure Synapse – You can import your Synapse model to make predictions on your customer data.

Responsible AI

Predictions offer capabilities that help organizations create better customer experiences, improve business capabilities, and increase revenue streams. AI results can have a real impact on people's lives. It's important that you keep this in mind the more you use AI in your solutions. Responsible AI is the practice of designing, developing, and deploying AI with good intentions to empower employees and businesses. It includes impacting customers and society and allowing companies to build trust and scale AI confidently.

We highly recommend that you balance the value of your prediction against the effect that it has on biases that might be introduced in an ethical manner. For more information about different techniques and processes for responsible machine learning, see Use custom models from Azure Machine Learning.

Machine learning concepts

Before you learn about how to create different machine-learning modules, you should learn about a few key concepts and terminology.

  • Workspace - Workspaces are centralized places to manage resources for training and deployment of models and to store assets that you create when using Azure Machine Learning.
  • Computes - Computes are any machine or set of machines that you use to run your training script or host your service deployment.
  • Datasets and datastores – Help make it easier for you to access and work with your data. By creating a dataset, you create a reference to the data source location along with a copy of its metadata.
  • Environments - Where training or scoring of your machine learning model happens. The environment specifies the Python packages, environment variables, and software settings around your training and scoring scripts.
  • Experiments - A grouping of many runs from a specified script.
  • Models - Pieces of code that produce output from an input. Creating a machine learning model involves selecting an algorithm, providing it with data, and tuning hyperparameters.

Azure Machine Learning studio

Azure Machine Learning is a cloud service for accelerating and managing the machine learning project life cycle. Machine learning professionals, data scientists, and engineers can use it in their daily workflows. You can import and use custom models that you create in Azure Machine Learning in Dynamics 365 Customer Insights - Data.

To get started using models, follow these steps:

  1. Set up an Azure Machine Learning workspace – You can create workspaces and access them in Azure Machine Learning studio.
  2. Create a batch pipeline with Azure Machine Learning designer - Azure Machine Learning designer provides a visual canvas where you can drag and drop datasets and modules. You can integrate a batch pipeline that you create from the designer into Customer Insights - Data if you configure them accordingly.
  3. Import a pipeline into Customer Insights – Data - The designer provides the Export Data module that allows the system to export the output of a pipeline to Azure storage. Currently, the module must use the datastore type Microsoft Azure Blob Storage and parameterize the Datastore and relative Path. Customer Insights - Data overrides these parameters during pipeline implementation with a data store and path that's accessible to the product.

Azure Machine Learning resources

For more information about Azure Machine Learning, select the following links: