Azure Databricks Released for Machine Learning

• Azure Databricks is an analytics platform for designing and deploying machine learning applications.
• The hosted service automates the configuration and scaling of servers and tools to simplify some of the complexity of building large-scale analytic applications.
• Integration with other Azure services can create a fully hosted solution.

Azure Databricks is a Platform as a Service offering that provide analytics, orchestration, and collaboration tools for creating and operating machine learning applications. It is based on Apache Spark, a popular Big Data machine learning package, but can also be used for general analytic workloads and custom applications that use Spark. The service became generally available in Apr. 2018, after a short preview period.

Service Overview

The service offers two deployment types, called clusters:

Data Engineering clusters are used for operational activities on predeveloped models. For example, a manufacturer might run automated batch processes to determine machine failure rates, either on a set schedule or activated by an event. The clusters are designed for transient processes and are controlled by scheduling tools that automatically create and delete the clusters on which it runs.