For years, Euronext FX's data warehouse did what it was designed to do. But as the business grew, so did the complexity and volume of data flowing through it, and several realities began to converge. Greenplum had moved to a closed-source model, meaning every upgrade, security patch, and new feature would be dependent on a single vendor, with support and licensing costs expected to climb accordingly. At the same time, the existing stack was approaching the limits of what it could deliver in terms of availability and scalability for a 24/5 marketplace with zero tolerance for disruption.

The conclusion was clear: staying the course would mean accepting growing risk, growing cost, and a ceiling on ambition.

Choosing sovereign infrastructure and the right partner

When evaluating alternatives, Euronext FX set a firm criterion: backward compatibility. The replacement had to be a true drop-in, something the team could test in staging and promote to production without data movement, schema changes, or any impact on client applications. After considering both commercial and open-source options, EDB Postgres AI (EDB PG AI) for WarehousePG stood out for two reasons: deep technical expertise and the kind of long-term stability that matters when choosing infrastructure partners in a high-stakes environment.

"EDB was critical here," Zeleniy explains. "They helped us understand the risk. They were the kind of vendor that was willing to go deeply technical with us, and that is something that ultimately sets you up for long-term success."

Months of careful planning preceded the actual migration, but when the cutover came, it happened in hours. For a marketplace that runs 24 hours a day, five days a week, the speed and smoothness of the transition was essential.

Support that goes beyond the first response

Since the migration, the capability that has stood out most for the Euronext FX team isn't a performance benchmark or a feature, but rather the nature of EDB's support model. In an industry where 24/7 support is common, the distinction EDB offers is the commitment to time-to-remedy, not just time-to-response.

"The commitment to remedy the issue within a short timeframe is harder to come by, and that’s critical to our business," Zeleniy says. For a platform where downtime has direct market consequences, that distinction carries real weight.

A sovereign data and AI foundation for what comes next

With a stable, open-source foundation in place, Euronext FX is now focused on what comes next. Zeleniy sees two forces reshaping what clients will expect: the move from batch processing to real-time streaming analytics, and a shift away from complex dashboards toward natural language access to data.

Conversational analytics—the ability for business users to query large, complex datasets by simply asking a question—is a high-priority area of exploration. The goal is to give customers instant, deep insight into market data without requiring data science expertise. The sovereign data foundation being built today, Zeleniy believes, is what will make that competitive advantage possible.

"We are turning from just a matching engine into an intelligent, data-driven trading ecosystem," he says. "The foundation that we're building today in data analytics will give us that core competitive advantage."