Enterprise AI deployments are stalling not because agents are hard to build, but because organizations lack the data infrastructure to run them reliably at scale. The shift from chatbots to autonomous, multi-step agents has exposed a structural gap in agentic AI development.

Oracle Corp. is positioning the database as the center of gravity for enterprise agentic AI, arguing that the future of intelligent applications will be determined not by model performance alone, but by how deeply AI is integrated with the underlying data layer. That conviction is now being translated into a concrete architecture, according to Tirthankar Lahiri (pictured), senior vice president for mission-critical data and AI engines at Oracle.

“Agentic systems are going to become the future of application development. They’re the present and they are the future,” Lahiri told theCUBE. “Many organizations are still struggling to realize value from agents, because ultimately, agents are only as good as their data.”

Lahiri spoke with theCUBE’s Dave Vellante at the Oracle Data Deep Dive NYC event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed Oracle’s strategy for embedding agentic AI development as well as how the company is tackling AI data security and open standards. (* Disclosure below.)

Agentic AI development anchored in unified memory and the data layer

Oracle’s approach challenges the prevailing assumption that agentic AI is primarily an orchestration problem. Rather than advocating for a separate agent layer sitting above fragmented data stores, Oracle is collapsing the stack — running agent logic as close to the data as possible. The company’s AI Database Private Agent Factory and Autonomous AI Vector Database reflect that thesis, giving developers and business users alike a low-friction path to build and deploy agents against live enterprise data without moving it between systems, Lahiri explained.

“There are basically two types of agents. There’s reasoning-centric agents and data-centric agents,” he said. “The data-centric agents are really best run co-located with data. We want to eliminate the need for multiple round trips — multiple database accesses. Architecting agentic processing along with data access avoids fragmented or fractured AI. You get AI that runs on clean, real-time, current data without the need to split your data in multiple repositories.”

Central to Oracle’s architecture is what the company calls Unified Memory Core — a capability that derives the full spectrum of agent memory constructs, from short-term context to long-term factual associations, from a single unified data store, Lahiri explained. Rather than routing agents to separate graph, document or vector databases for different reasoning tasks, Oracle lets a single underlying data layer answer all of those needs simultaneously. This agentic AI development eliminates the synchronization overhead and consistency risks that come with managing multiple specialized systems.

“Sometimes you want associations and you want a knowledge graph. Sometimes you just want a factual representation of an event that happened,” he said. “That derivation, if it’s done in place with the actual data, is current, it’s consistent and it’s fully secure. We call that the Unified Memory Core for that reason, which is much more efficient than using multiple storage systems to represent the different kinds of memory.”

The same data-proximity logic extends to Oracle’s approach to AI data security. As agents move from answering questions to taking action — executing transactions, accessing sensitive records, running business processes — security enforced at the application layer becomes inadequate. Oracle’s answer is what it calls Deep Data Security: policy enforcement embedded directly in the database, ensuring that even a dynamically generated or adversarially injected query cannot return data the authenticated user is not authorized to see.

“The problem we have today is in many systems, security is built in the application tier,” Lahiri said. “The only way to solve this problem is securing data at the source. Even if the query is malformed, it can’t return data it shouldn’t show. That’s really what deep data security gives you — and I think in this AI world, that’s the only way to secure data.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the Oracle Data Deep Dive NYC event:

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

Photo: SiliconANGLE