migrate-rebuild-integrate-with-client-app.md

title	description	services	ms.service	ms.subservice	ms.topic	ms.custom	author	ms.author	ms.date
Migrate to Azure Machine Learning - Consume pipeline endpoints	Learn how to integrate pipeline endpoints with client applications in Azure Machine Learning as part of migrating from Machine Learning Studio (Classic).	machine-learning	machine-learning	studio-classic	how-to	kr2b-contr-experiment	xiaoharper	zhanxia	05/31/2022

Consume pipeline endpoints from client applications

[!INCLUDE ML Studio (classic) retirement]

In this article, you learn how to integrate client applications with Azure Machine Learning endpoints. For more information on writing application code, see Consume an Azure Machine Learning endpoint.

This article is part of the ML Studio (classic) to Azure Machine Learning migration series. For more information on migrating to Azure Machine Learning, see the migration overview article.

Prerequisites

• An Azure account with an active subscription. Create an account for free.
• An Azure Machine Learning workspace. Create an Azure Machine Learning workspace.
• An Azure Machine Learning real-time endpoint or pipeline endpoint.

Consume a real-time endpoint

If you deployed your model as a real-time endpoint, you can find its REST endpoint, and pre-generated consumption code in C#, Python, and R:

1. Go to Azure Machine Learning studio (ml.azure.com).
2. Go the Endpoints tab.
3. Select your real-time endpoint.
4. Select Consume.

Note

You can also find the Swagger specification for your endpoint in the Details tab. Use the Swagger definition to understand your endpoint schema. For more information on Swagger definition, see Swagger official documentation.

Consume a pipeline endpoint

There are two ways to consume a pipeline endpoint:

• REST API calls
• Integration with Azure Data Factory

Use REST API calls

Call the REST endpoint from your client application. You can use the Swagger specification for your endpoint to understand its schema:

1. Go to Azure Machine Learning studio (ml.azure.com).
2. Go the Endpoints tab.
3. Select Pipeline endpoints.
4. Select your pipeline endpoint.
5. In the Pipeline endpoint overview pane, select the link under REST endpoint documentation.

Use Azure Data Factory

You can call your Azure Machine Learning pipeline as a step in an Azure Data Factory pipeline. For more information, see Execute Azure Machine Learning pipelines in Azure Data Factory.

Next steps

In this article, you learned how to find schema and sample code for your pipeline endpoints. For more information on consuming endpoints from the client application, see Consume an Azure Machine Learning endpoint.

See the rest of the articles in the Azure Machine Learning migration series:

• Migration overview.
• Migrate dataset.
• Rebuild a Studio (classic) training pipeline.
• Rebuild a Studio (classic) web service.
• Migrate Execute R Script.