Companies turn to EnterpriseDB for AI data control

Organisations are increasing their use of EnterpriseDB's EDB Postgres AI platform, a trend the company links to demand for greater control over data used in AI systems.

Banks, insurers, retailers and trading businesses are using the platform for core systems, analytics and AI workloads. EnterpriseDB argues that as AI projects move from early trials into live operations, companies are placing more emphasis on data sovereignty.

Research by MIT Technology Review Insights, produced in partnership with EnterpriseDB, found that commitment to AI and data sovereignty was the strongest predictor of AI success, according to the company. Organisations most focused on controlling their data and infrastructure delivered five times the return on investment of peers, it said.

Nancy Hensley, Chief Product Officer at EDB, said the growing use of autonomous AI systems is shifting attention to the data layer behind them. "The agentic era is raising those stakes sharply: As autonomous agents move into production, they act on enterprise data at a speed and concurrency that turns the data layer into the control point for the entire AI strategy-and makes owning and governing it a source of capability, not just protection. It's why the world's most demanding enterprises are building that foundation on EDB PG AI," Hensley said.

Korean migrations

Two of the examples highlighted by EnterpriseDB came from South Korea's financial sector, where database migrations have been driven by cost, licensing and operational flexibility.

Industrial Bank of Korea, a lender focused on small and medium-sized businesses, moved 15 core systems to EDB Postgres AI. The bank said its previous proprietary database limited customisation under restrictive licensing terms, while the new platform allowed SQL, procedures and packages to be converted with limited changes.

The bank also pointed to lower software spending and the option to extend the same database base into AI services. "We've achieved a significant reduction in licensing costs compared to Oracle, and that's a core win that resonates with leadership on the IT budget side," said Park Cheol-min, Manager, IT Operations Division, IBK. "Being PostgreSQL-based, EPAS supports extensions like pgvector for AI and vector operations. That gave us a platform with the scalability to extend into AI services down the road. Beyond simple data management, we'll keep advancing our open source, DB-centered data infrastructure strategy," Park said.

Shinhan EZ Insurance, a digital insurer, also rebuilt its core architecture on the platform after moving its full core system to the public cloud. The insurer said its legacy database licensing model did not fit the flexible scaling expected in cloud infrastructure and that maintenance costs had risen.

For the insurer, the migration was tied to service continuity as well as future technology choices. "Migrating our core systems to the public cloud was an enormous challenge in the financial sector. EDB's professional support structure, backed by a global engineering team, played a decisive role in securing our operational stability," said Jin-sun Kim, Infrastructure Lead, IT Planning Team, Shinhan. "In a financial system, 24/7 uninterrupted service is simply nonnegotiable, and our greatest achievement is the ability to instantly support new services requested by business teams. We'll keep embracing the latest technologies, from AI to big data," Kim said.

Broader customer base

Outside financial services, EnterpriseDB cited use of the platform by advertising technology company MNTN, trading venue Euronext FX and South Korean retailer Kyobo Book Centre. These users adopted the software to reduce dependence on specific vendors, manage large-scale data workloads and gain more direct control over compliance and operating costs, the company said.

MNTN uses EDB Postgres AI for ad-tech workloads that require large-scale analytical processing, according to EnterpriseDB. Euronext FX deployed it across four data centres, while Kyobo Book Centre adopted it after moving away from a 50 TB cloud-only data warehouse that it viewed as too costly.

A common theme across these deployments is the use of one Postgres-based platform for transactional processing, analytics and AI-related work, EnterpriseDB said. It argues that this reflects a wider industry effort to reduce the complexity and cost of maintaining separate systems for live operations and data analysis.

Hensley said the issue becomes more acute as AI systems interact directly with operational data. "That convergence is exactly what the agentic era requires. Because AI agents retrieve, analyze, decide, and act against live enterprise data in continuous, high-volume workflows, they amplify the cost and latency penalties of stitching together separate platforms for transactions, analytics, and AI," Hensley said.

EnterpriseDB also pointed to recent industry awards for data management and open source community work as evidence of its standing in the market. It said the platform is being adopted by organisations seeking greater control over where data sits, how it is governed and which infrastructure it runs on.

Among the examples cited, Euronext FX eliminated vendor risk and regained open source control across four data centres worldwide, according to EnterpriseDB.