While the hype around AI keeps growing louder, Igor Kersic is quietly focused on something more foundational: making AI work in real time. As CTO of Ververica, the original creators of Apache Flink, Kersic isn’t chasing headlines. He’s helping power the data infrastructure that underpins next-gen applications, from fraud detection at scale to dynamic risk monitoring in finance.
The shift has been years in the making. Ververica’s roots trace back to Berlin, where the company built its streaming data platform on top of Flink, designed for speed and adaptability. But today, that platform is fueling a much larger movement. Companies like Booking.com use Ververica to process massive amounts of real-time data to detect fraud as it happens, not hours or days later. The same is true in banking, where regulators are tightening expectations around anti-money laundering and real-time risk detection. Mandates like BCBS 239 and BCBS 230 are forcing enterprises to rethink how quickly their systems can respond.
And that’s where Ververica sits, in the architectural core of those real-time workflows. In Kersic’s view, legacy batch processing still dominates much of the data world. Most people assume their AI tools are operating in real time, but in practice, the freshness of that data is lagging. What’s needed is a “streaming-first” mindset, where data flows are continuous, not staged. Ververica’s technology acts as a bridge, pushing organizations from reactive data systems into proactive, predictive ones.
Kersic is also clear about what real-time AI isn’t. It’s not about replacing humans, but augmenting them. Whether it’s within enterprise software, fraud detection tools, or AI agent frameworks, Ververica is betting on a future where AI and humans collaborate fluidly. Their roadmap includes flexible deployment across on-premise, cloud, and air-gapped environments, as well as open-source contributions to Flink agents — tools that bring AI context and adaptability to edge applications.
The tech is moving fast, but Kersic is pragmatic. He’s been in engineering for more than two decades, and he understands the push and pull between innovation and enterprise reliability. Clients in banking and fintech aren’t chasing what’s flashy. They need security, scale, and compliance to work in harmony with new tools. That balance, he admits, is one of the toughest parts of the job. But it’s also what makes the work meaningful.
At its core, Kersic’s approach is human. He believes technology only matters if it solves real-world problems, and that AI should serve people, not distance them. Whether it’s enabling faster financial decisions or helping reduce fraud risk for global marketplaces, the goal is always grounded: build tech that makes life easier, safer, and smarter.
In the evolving ecosystem of real-time AI, Ververica is one of the quiet forces making the future possible, one data stream at a time.
Want more Plugged In? Read more or listen on Apple Podcasts.
