Database Modernization: Oracle’s Path to AI at Scale

The motivation behind database modernization has traditionally been greater efficiency and cost savings. However, theCUBE Research also found that there is a larger prize at stake: AI projects built on a modernized foundation can generate significant value.

The ultimate goal is to overcome fragmented data that can slow down teams and complicate governance. AI projects need to be repeatable and scalable in production, and this will be a key focus of discussion at Oracle AI World in Las Vegas later this month.

“The strategic question for Oracle customers is: Are we simply reducing infrastructure costs or are we creating a better platform for repeatable AI execution?” said Dave Vellante of theCUBE Research. “The value is not just in automation, standardization and operational efficiency. The greater opportunity is in leveraging this foundation to move earlier into the AI ​​learning curve, freeing skilled teams from routine administration, improving governance consistency and accelerating the development of AI use cases over time. Database modernization is still an IT decision. But increasingly it is also an AI readiness decision that impacts every part of the business.”

TheCUBE, SiliconANGLE Media’s livestreaming studio, will be covering the latest news and announcements from Oracle AI World on October 28 in Las Vegas. Coverage includes on-site reports and exclusive interviews with Oracle executives and industry leaders as theCUBE examines how Oracle integrates agentic AI into enterprise operations, connects trusted data with intelligent automation, and meets the infrastructure, security and governance needs of AI at scale. (*Disclosure below.)

Database modernization to ensure trust

The upcoming Oracle event will provide the company with an opportunity to showcase recent advances in a number of key product areas. One of them concerns the implementation of guardrails and trust.

AI-generated solutions must be accurate, secure and reliable so that companies can trust their results. According to Juan Loaiza, executive vice president of mission-critical database technologies at Oracle, this has become essential as autonomous agents build applications and take actions.

“Guardrails are not enough. You can’t just put AI on an architecture that was never designed for these challenges that AI presents and then expect it to work,” Loaiza said during a pre-event conversation with theCUBE. “What we do is different. We ensure trust across the entire technology stack. The only way we can build AI solutions we can trust is through common architecture – AI, data and AppDev in one.”

Earlier this year, Oracle unveiled its three-step strategy to address the security challenges posed by the broader use of artificial intelligence. AI systems have created new pathways to sensitive corporate data, and Oracle has implemented a range of tools focused on database security, patching, testing and lifecycle management.

“There is so much optimism around AI and how it can improve productivity, but there is also a lot of concern when it comes to cyberattacks,” said Arpan Shah, senior VP of Oracle Database at Oracle, in a recent interview with theCUBE. “One of the key focuses for us in Las Vegas is to really demonstrate what we do in the product, how we design features that really start at the data level and go all the way to the app level to help customers move from AI prototypes to trusted AI apps at scale.”

Improvements for Lakehouse and AI database technologies

Another focus is on intelligent data architecture. Oracle’s Autonomous AI Lakehouse combines the company’s Autonomous Database capabilities with Apache Iceberg’s open table format to provide an interoperable foundation for analytics that spans all major clouds.

According to Nipun Agarwal, SVP of Oracle Autonomous AI Lakehouse at Oracle, Oracle will introduce a number of new features in Autonomous AI Lakehouse.

“The Lakehouse is transforming from an analytics platform to a decision-making backbone for enterprise AI,” Agarwal told theCUBE. “What this essentially means is that in the past, customers came to a lakehouse to run their analytics workloads. But now Lakehouse is the right, suitable platform for both analytics and AI workloads.”

Oracle’s Autonomous AI Database also provides a fully automated, cloud-based data platform that leverages machine learning and artificial intelligence to manage, optimize, patch and secure itself. Its multicloud capabilities give organizations the choice of running on Oracle Cloud Infrastructure or on Oracle-managed services on Amazon Web Services Inc., Microsoft Azure and Google Cloud. The database provides multimodal data management, supporting transactional and analytical processing while helping to standardize and automate database operations.

“Of course, at AI World we will be talking a lot about building and integrating AI agents with Autonomous AI Database,” said Çetin Özbütün, EVP of Autonomous AI Database technologies at Oracle, during an appearance on theCUBE. “But if you only look at how LLMs interact with enterprise data from an application perspective, you miss a big part of the picture. An often overlooked factor is that AI applications need to be built on the right data foundation. And this is where the true strengths of the Autonomous AI Database come into play.”

New platform features

Oracle’s platform capabilities continued to improve this year with the introduction of features such as APEXlang. The human-readable application definition language provides leading programming agents with a structured application goal, while Oracle manages runtime, security and application lifecycle.

“What APEXlang does is it takes all of these configurations or all of these properties that with a low-code system you traditionally define through a user interface, but they are captured in a formal language with a formal grammar and can be compiled,” said Mike Hichwa, SVP of software engineering at Oracle, during a recent conversation with theCUBE. “So you can use a coding agent like Codex or Claude and just say, ‘Build me an application to do something.’ And it will produce this language.”

Oracle has gradually expanded its highly available Exadata service to target AI and location-sensitive use cases. However, live migration of database VMs with Remote Direct Memory Access in Exadata was nearly impossible.

In June, Oracle announced Exadata VM Live Migration, a feature that allows a live Oracle database to move between hosts while continuing to work on Exadata. According to Jia Shi, group vice president of software development at Oracle, in an interview with theCUBE, this new feature will be described in more detail during Oracle AI World.

“Before this June, no one could live move their database VM running RDMA from one host to another,” Shi said. “That’s why we’re very excited to announce that we’ve finally solved this technical challenge. This is truly unique in the sense that we can introduce the feature on Exadata that allows customers to move their database VM running an active database workload from one server to another with all RDMA connections intact.”

TheCUBE event live stream

Don’t miss theCUBE’s coverage of Oracle AI World on October 28th. You’ll also be able to access theCUBE’s exclusive content on-demand after the event.

How to watch theCUBE interviews

We offer several ways for you to watch theCUBE’s coverage of Oracle AI World, including theCUBE’s dedicated website and YouTube channel. You can also find all of our coverage of this year’s events on SiliconANGLE.

TheCUBE Podcasts

SiliconANGLE’s theCUBE Pod is available on Apple Podcasts, Spotify, and YouTube so you can enjoy it on the go. In each podcast, SiliconANGLE’s John Furrier and Dave Vellante discuss the biggest trends in enterprise technology, including AI, cloud, cybersecurity and infrastructure, with context and analysis.

SiliconANGLE also produces our weekly Breaking Analysis program, in which Vellante examines key developments in enterprise technology and combines insights from theCUBE with spending data from Enterprise Technology Research, available on Apple Podcasts, Spotify and YouTube.

guests

At Oracle AI World on October 28, theCUBE will speak with Oracle executives and industry managers about integrating agent AI into enterprise operations, connecting trusted data with intelligent automation, and scalably addressing AI’s infrastructure, security and governance needs.

(*Disclosure: TheCUBE is a paid media partner for Oracle AI World. Neither Oracle, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over theCUBE or SiliconANGLE content.)

Image: SiliconANGLE