Oracle pitches autonomous capabilities for entire cloud platform
AI techniques in PaaS app development services, data integration, analytics and security to cut costsPrint
13 February 2018 | 0
Oracle is raising its bid to attract enterprises moving to the cloud by extending AI-based automation throughout its platform as a service (PaaS) offerings, claiming that the new automated services will cut costs and reduce management headaches.
The company is using artificial intelligence (AI) techniques, including machine learning (ML), to enable the various services in its Oracle Cloud Platform to do self-maintenance, such as optimising workloads and installing patches.
The basic pitch to CIOs is straightforward. “It’s really reducing costs and increasing productivity — that’s what we’re aiming for, where CIOs can really take care of focusing their resources and investment on strategic initiatives and have the services run by Oracle be very self-running,” said Amit Zavery, executive vice president of product development, Oracle Cloud Platform. “We want to provide autonomous capabilities to eliminate the human labour associated with provisioning, upgrades, backup, recovery, and troubleshooting,” Zavery said.
Automation as a differentiator
This sort of automation is key to how Oracle is trying to differentiate itself from competitors like Amazon and Microsoft that have a longer history in the cloud, not to mention much bigger market share.
Last October at its OpenWorld conference, Oracle announced its Autonomous Database Cloud, powered by Oracle Database 18c. Over the next few months, the self-tuning features of 18c will be brought across Oracle’s entire cloud platform, the company announced at its CloudWorld event in New York.
To a certain extent, users have for some time expected cloud providers to do some patching and provisioning as part of basic services. But Zavery points out that PaaS services in particular still require a lot of human intervention and manual management. “A lot of patching doesn’t happen automatically and a lot of systems don’t get upgraded regularly,” Zavery said.
Meanwhile, machine learning and various automated services are being added to enhance specific functions provided by Oracle’s PaaS applications development, application and data integration, analytics and security services.
AI for chatbots, data integration, analytics and security
Application development services will include chatbot technology designed to help developers apply chatbots to their applications. Oracle has modelled a variety of different chatbot use cases so developers don’t have to start from scratch, and has incorporated machine learning into the software so that the chatbots, through usage, get better at understanding what users intend to do on a website. Developer services also include automated identification and remediation of security issues in code.
New application and data integration capabilities have been designed to automate the flow of data, with data prep pipeline and data lake creation. Oracle has a good understanding of what database schemas and structures look like and also has gained a lot of knowledge from custom integrations it has done, Zavery said. “Once a user defines the kinds of things he wants to integrate, we can take over connecting systems, doing the mapping and providing endpoint connectivity,” he said.
For analytics services, a machine-learning based data preparation engine cleans up data and prepares it in a way that can be used in data lakes and for data discovery. Though Oracle has deep roots in relational databases, the company offers a NoSQL database and also Hadoop and Spark services for unstructured data.
New security and management controls use machine learning to, among other things, understand user behaviour and intent to prevent and intercept data leaks across structured and unstructured data repositories.
Oracle is highly motivated to do everything it can to differentiate itself from cloud leaders like Amazon, Microsoft and Google. On the PaaS side, strong competitors also include Salesforce. Enterprise customers are flocking to the cloud, and less apt to sink money into traditional on-premises databases and applications.
IaaS, PaaS growth, but clouds on the horizon
In fact, in its last quarterly earnings report, for the period ending 30 November, Oracle said revenue for new on-premises software licenses was flat year over year. The good news for the company was that total cloud revenue was up 44% and IaaS and PaaS revenue together, were up 21%. (Oracle doesn’t break out PaaS and IaaS sales separately.) In absolute terms, though, Oracle’s quarterly on-premises revenue still dwarfed cloud revenue — $7.8 billion (€6.3 billion) compared to $1.5 billion (€1.2 billion).
At the moment, much of Oracle’s appeal in the cloud is for current on-premises customers who want to retain their investment in Oracle software as they move to hybrid systems and cloud technology. But for Oracle, that might not be enough to keep up a healthy growth rate.
Oracle forecast cloud revenue for the current quarter would grow 21% to 25%, disappointing analysts, who had expected growth would be closer to 30%.
Though Oracle has a huge installed base of relational database users, it is going to have to start winning over more new customers for cloud offerings.
Look for Oracle to continue to play up AI and automation as it strives to win over new users to its cloud services.
Meanwhile, Oracle also said that it will be opening 12 new data centre regions. Locations in Asia and the Middle East include China, India, Japan, Saudi Arabia, Singapore, and South Korea. Locations in Europe include data centres in The Netherlands and Switzerland. In North America, new locations include two in Canada and two new US locations to support US Department of Defence workloads.
IDG News Service