Oracle updates cloud data warehouse to boost access for data analysts

Oracle today updated its Autonomous Data Warehouse to enable data analysts to load, transform, and generate insights from data with no intervention on the part of an internal IT team required.

The latest update to the Oracle data warehouse cloud service also enables data analysts to automatically create business models and discover patterns, along with providing a set of tools for preparing data and building machine learning models guided by AutoML, a set of open methods and processes for building AI models.

Other capabilities that have been added include support for the Python programming language, cognitive text analytics, graphs that can be invoked using a set of visualization tools, and an ability to deploy and manage native in-database models and ONNX-format classification and regression models outside of the core database.

The goal is to make it simpler for both professional and citizen data analysts to access data whenever they want using a fully autonomous platform, said George Lumpkin, VP of product management for Oracle. “We’re trying to provide what a cloud data warehouse should be,” he said.

As the provider of a data warehouse that is widely employed in the enterprise, Oracle is attempting to fend off increased competition from cloud service providers such as Amazon Web Services (AWS), Microsoft, Google, and Snowflake. Oracle alternatively makes available data warehouse platforms from a single vendor that can be deployed in both its cloud and on-premises IT environments at a time when the bulk of enterprise data continues to reside in local datacenters. Rather than requiring organizations to make a wholesale shift to the cloud, Oracle enables them to make that transition at their own pace, Lumpkin said.

In contrast to rival cloud data warehouses, Oracle has built its approach around a managed service that eliminates the need to dedicate IT professionals to managing, securing, and maintaining a cloud platform, Lumpkin added.

Oracle also provides access to a low-code Oracle APEX (Application Express) tool that makes it possible for both “citizen integrators” and professional developers to build applications that can be deployed via REST application programming interfaces (APIs), Lumpkin added.

It’s too early to say to what degree business units within organizations might be willing to enable data analysts and scientists to access, manage, and analyze data without any oversight from a central IT function. However, as IT becomes increasingly automated, it’s apparent that many of the manual data management tasks that used to require an IT professional are falling by the wayside. The time when end users had to wait days for an IT professional to set up am SQL query to generate a report is all but over as self-service tools become more widely available.

In effect, Oracle is making a case for transferring data management tasks to its platform. Less clear is to what degree that might lower the total cost of IT for companies. It’s worth remembering that as data becomes more accessible, usage will increase so organizations may wind up spending more on analytics. The difference is that they will hopefully be able to derive more business value from data that has become easier to interrogate in near real time.

In the meantime, the odds that most organizations will migrate all their relevant data to the cloud anytime soon is low. In fact, most organizations will be managing multiple data warehouses for years to come. The challenge will be determining what type of data needs to reside where based on use cases that are becoming more varied with each passing day.