Data managers are on the hook to deliver AI and intelligent agents to the enterprise when needed. To achieve these capabilities, they are working within diverse and decentralized data environments while addressing challenges by building AI-powered systems and applications.
These are some of the takeaways from a new survey of 218 data managers, developed and fielded by Unisphere Research in partnership with Rimini Street (“Oracle Database Strategies and Support Survey—The Diversification and Decentralization Revolution,” Unisphere Research).
Today’s data teams work within a diverse ecosystem of database tools. While the survey focused exclusively on Oracle users, these users also oversee a range of database brands, including open source and cloud databases. A majority (59%) also use SQL Server; 45% also use MySQL; 40% use PostgreSQL; and 28% use Amazon RDS. Another 22% have MongoDB, and 19% have IBM DB2 databases.
The findings of this survey reflect this evolution: Organizations are experimenting with diverse database ecosystems, but the true revolution is happening at the application layer. Survey respondents report a strong appetite for AI and machine learning (ML) capabilities: More than half are seeking closer integration between their databases and popular AI/ML frameworks, and nearly half want native support for large language models.
This demand for advanced application features is driving organizations to explore new database platforms, cloud-native solutions, and multi-model architectures.
The survey also reveals that most organizations are not abandoning their existing databases outright but are instead expanding their ecosystems to support new application requirements. This is evidenced by the widespread use of multiple database types, including relational, NoSQL, and cloud-native options. The move toward decentralization, cloud, and AI integration is a direct response to the evolving demands of business applications.
The survey also revealed that there aren’t enough skilled people to manage AI/ML initiatives (cited by 52%), and many (24%) are still falling short with basic database skills.
A majority cite the increasing technical complexity of tools, platforms, and applications, as well as the continuing rise of AI or ML initiatives within the enterprise. The greater variety of database platforms to manage also is adding to the acuity of the data skills shortfall.
Database performance issues continue to take a major bite out of organizational time and resources, the survey also reveals. Many respondents report being impacted by database performance issues on a monthly basis or more.
Database performance issues affect more than strictly data operations. Slow response times or glitches can cut into employee productivity, as well as an organization’s revenues and market position.
These issues continue to take a bite out of organizational time and resources, the survey shows. Most respondents (62%) report they are impacted by database performance issues on a monthly basis or more. Notably, only 15% report they were “rarely” impacted by database performance issues.
With issues that affect organizational success arising on a weekly, and sometimes daily, basis for more than 1 in 4 respondents, it’s no surprise that addressing performance issues ranks as the top time-consuming task faced by database teams.
Survey results show organizations are rebalancing their database environments. At present, 45% of databases are hosted on-premises, 18% in co-located data centers, and 37% on cloud infrastructure.
Within the next 3–5 years, cloud-hosted databases are expected to rise to 53%, while on-premises hosting will decrease to 30%, and co-location will remain stable at 17%.
This transition reflects a pragmatic approach: Most enterprises continue to operate hybrid environments, expanding their options and balancing flexibility, control, and cost. Decisions about hosting are shaped by business needs, application requirements, and operational priorities—not by a universal shift to managed database services.