New Oracle AI database targets cost, data residency concerns
The tech giant's latest offering is a scaled-down version of its Exadata platform and helps differentiate its lineup of databases from those of competing providers.
Oracle is taking aim at rising concerns over cost control and data residency with its latest AI database.
Introduced on July 22, Oracle Base Database Cloud@Customer is a scaled down version of Oracle's Exadata Cloud@Customer database. The new offering is purpose-built for small and mid-sized data workloads, so customers don't have to use the compute power of databases optimized for large-scale workloads.
In addition, the new AI database is designed to enable Oracle customers to run their data workloads in their locations of choice.
Driven by AI development, which requires far more data than analytics or traditional machine learning projects, the cost of data management workloads is on the rise. Meanwhile, as enterprises build agents and other AI applications that autonomously access and replicate data across multiple regions, data residency concerns are growing.
Given that Base Database Cloud@Customer addresses both, it is a valuable addition for existing Oracle AI Database users as well as potential new ones that don't need to run massive-scale workloads, according to Holger Mueller, an analyst at Constellation Research.
"It is a key step to run a less complex version of the Oracle Database," he told TechTarget. "It creates value for those customers who want or need to run things on premises. Simpler is also a key value proposition for small and mid-sized businesses. So, all in all, a very good move."
Stephen Catanzano, an analyst at Omdia, a division of TechTarget, similarly noted the value of the newest addition to Oracle's lineup of AI Databases.
"Oracle Base Database Cloud@Customer is a solid addition because it extends Oracle's cloud operating model to mid-sized workloads that previously had no good hybrid option," he told TechTarget.
Previously, potential customers with data residency requirements or latency-sensitive applications at remote sites were forced to use traditional on-premises deployments or databases such as Exadata Cloud@Customer that were more powerful -- and expensive -- than they needed, Catanzano continued.
"This matters because it democratizes hybrid cloud access for smaller organizations, departments, and remote facilities that need to keep data local but want cloud benefits like automated management, pay-per-use pricing, and simplified operations," he said.
Addressing user needs
Cost control is a continuing concern for many enterprises as they increase their investments in AI development and move more of their operations to the cloud, with Gartner finding that managing the expense of AI development limits what some organizations can build.
In response, some data management vendors are introducing performance improvements, new pricing models and other features designed to help customers curb some of their spending.
It is a key step to run a less complex version of the Oracle Database. It creates value for those customers who want or need to run things on premises. Simpler is also a key value proposition for small and mid-sized businesses. So, all in all, a very good move.
Holger MuellerAnalyst, Constellation Research
For example, AWS in December 2025 unveiled new pricing for its databases and capabilities that lowered the cost of storing and searching vectors. Similarly, Aerospike, Neo4j, Oracle and SingleStore have all made performance improvements to their databases to improve their efficiency.
Now, Oracle is further addressing cost control with Base Database Cloud@Customer in a move motivated by customer feedback, according to Ashish Ray, Oracle's senior vice president of database product management.
"Enterprise customers of Exadata Cloud@Customer who rely on it for their mission-critical, large-scale datacenter workloads have been looking for a smaller form factor Cloud@Customer product for their satellite locations, warehouses and regional offices," he said.
Growing concerns over data residency were an additional factor for Oracle when developing latest AI database, Ray continued.
"Organizations are facing data residency mandates and want the same cloud control plane … and highly efficient storage technologies to deploy their private AI workloads anywhere they require," he said.
Base Database Cloud@Customer is an AI database and hybrid cloud tool that runs on Oracle Data Infrastructure Cloud@Customer X11, a managed hybrid cloud infrastructure accessed through a simple subscription-based pricing model. The infrastructure includes two servers, so workloads can continue running despite unplanned power outages and planned maintenance.
In addition, because Base Database Cloud@Customer runs on the same infrastructure as other Oracle AI capabilities, it can be used in conjunction with Oracle AI Database 26ai and Oracle Database 19c, as well as AI development tools such as the following:
AI Vector Search so users can operationalize both structured and unstructured data for AI.
AI Database Private Agent Factory to easily build and deploy agents.
Private AI Services Container to secure data and enforce data residency requirements.
The result is that Base Database Cloud@Customer users can deploy complete agentic workloads in controlled environments.
"What stands out is the unified infrastructure that lets customers run databases, applications and agents on the same hardware behind their firewall," Catanzano said. "The ability to deploy complete agentic AI workloads with private data entirely at customer locations addresses a critical gap in the market, where organizations want AI capabilities but cannot send sensitive data to external cloud providers."
Matt Aslett, an analyst at ISG Software Research, similarly noted that the biggest benefit of adding Base Database Cloud@Customer isn't necessarily the issues it addresses -- cost and data residency -- but instead that it gives Oracle users another option when selecting an appropriate database for their work.
"The new product expands the breadth of scalability options available for customers deploying data and AI workloads in hybrid and on-premises environments, … without having to commit to the scalability provided by Oracle Autonomous Database on Exadata Cloud@Customer," he told TechTarget.
Setting the pace
Beyond the benefit to users, Base Database Cloud@Customer is evidence of Oracle's innovation, according to Mueller.
He noted that competitors, including Microsoft and IBM -- among others -- offer on-premises database capabilities. But those on-premises databases are designed for large-scale workloads, like most cloud databases now optimizing for AI rather than scaled down versions of a more powerful offering.
Meanwhile, other Oracle databases also help differentiate the tech giant from its peers, Mueller continued.
"Oracle is leading," Mueller said. "Oracle Autonomous Database is eight years old, and the equivalent initial offerings of, for example, IBM and Microsoft are only coming out now."
Catanzano likewise noted that Base Database Cloud@Customer is different than other database offerings.
He pointed out that other vendors offer public cloud database services or on-premises offerings, but Oracle is unique by providing integrated hybrid cloud database capabilities with a fully managed infrastructure. For example, Microsoft offers Azure SQL and AWS provides RDS, both of which are hybrid cloud databases that compete with Oracle's offerings, but neither integrates with AI capabilities running on the same infrastructure.
"This matters because the market is fragmenting between cloud-first vendors and traditional database companies, and Oracle is carving out a middle path that acknowledges real-world constraints while delivering genuine cloud benefits," Catanzano said.
While Oracle may have certain competitive advantages, most data and AI providers are now addressing data residency requirements by providing numerous deployment choices, according to Aslett. However, Oracle continues to distinguish itself with its level of integration, he continued.
"Oracle has a distinct advantage over many of its software-focused competitors by virtue of its ability to address the full combination of AI and data requirements through integrated AI and data platforms, as well as its investment in Cloud@Customer infrastructure," Aslett said.
Looking ahead
As Oracle plots its AI Database strategy for the second half of 2026, data security features prominently, according to Ray.
Agentic AI is leading to new threats by enabling attackers to discover vulnerabilities and automate attacks at exponentially greater speed than in the past. In addition, because agents autonomously access data at greater speed and scale than humans, they have the potential to expose security gaps or create new ones.
"As AI accelerates the need for stronger security, Oracle remains focused on helping organizations secure data at the source, secure data at speed and secure data through resilience," Ray said. "By making it easier for organizations to assess, patch, upgrade and protect their databases, Oracle is helping them build a secure foundation for the next generation of AI-driven applications."
Mueller noted that Oracle would be wise to focus on improving the vector embedding capabilities in its AI Data Lakehouse that are often used in conjunction with transactional databases.
"It's not really new, but [Oracle has] been relatively quiet on the topic," he said.
Catanzano, meanwhile, suggested that Oracle expand its ecosystem of prebuilt AI agents and industry-specific agentic AI development templates in Private Agent Factory, so users can more easily derive value from their investments in AI.
"They could also introduce more granular consumption tiers and multi-site orchestration capabilities that let organizations manage fleets of hybrid cloud deployments across dozens or hundreds of locations from a central control plane," he said. "The current launch addresses the technical foundation, but the next wave of adoption will come from reducing time-to-value and operational overhead at scale."
Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than three decades of experience. He covers analytics and data management.
