• Home
• Fin tech

Financial technology companies must establish solid technical foundations to safely deploy Agentic Artificial Intelligence (Agentic AI) at scale, according to industry leaders speaking at an event hosted by FinTech Hive at the DIFC Innovation Hub. As Agentic AI rapidly transitions from experimental proof-of-concepts (POCs) to live production environments, building resilient infrastructure is becoming essential for firms aiming to maintain a competitive edge.

Titled "From POC to Production: Building the Foundations of Agentic AI in MENA FinTech," the event brought together executives and tech leaders from JetBrains MENA, Tabby, and Google for Developers. Panelists addressed key operational challenges in scaling Agentic AI, including modern engineering workflows, evaluation mechanisms, and managing the transition from developer-assist tools to human-supervised autonomous systems.

Accelerated AI Adoption in the UAE

The event comes as the UAE continues to consolidate its position as a global leader in artificial intelligence, supported by proactive regulatory frameworks designed to accelerate AI integration across both the public and private sectors.

A recent AI survey published by the Dubai Financial Services Authority (DFSA) revealed that more than half of licensed entities within the DIFC now utilize AI technologies, compared to roughly one-third in the previous year. Furthermore, the deployment of Generative AI tools nearly tripled within the same timeframe.

Speakers emphasized that the successful adoption of autonomous AI systems relies on more than just powerful foundational models—it requires advanced technical infrastructure, professional developer tools, and keeping human expertise directly in the loop to maintain strict governance and accountability.

Human Accountability in Autonomous Systems

Nadia Rinsky, Head of Market Development for MENA at JetBrains, highlighted that while the UAE offers an ideal regulatory environment for financial innovation, scaling Agentic AI presents a fundamental operational reality: accountability cannot be offloaded to a machine.

"AI cannot take responsibility when things go wrong. If a mission-critical financial system experiences an issue in production, the human developer remains fundamentally accountable for those systems," Rinsky noted.

She explained that the developer's role is evolving from manually writing code to orchestrating and auditing fleets of AI agents. She stressed the importance of equipping software engineers with professional-grade tools that instill complete confidence in the code integrated into live financial systems.