At the 22nd China International Finance Forum held in Shanghai on December 19-20, Gong Xiaojun, Chief Information Officer and Head of Technology & Operations at DBS Bank (China), delivered a keynote speech on building an intelligent financial ecosystem in the digital economy era.

Gong shared insights into DBS Group's Gen-AI transformation journey, which began over two years ago, highlighting several practical applications. The bank is reimagining the future of call centers by deploying intelligent AI agents that automatically identify and extract key customer information during calls, then relay structured data in real time to human agents. This model eliminates the need for agents to spend significant time on manual data entry and preliminary assessments, significantly improving efficiency. More importantly, it enables agents to transition from traditional service roles to higher-value sales or advisory positions.

"If this model is scaled in the future, it could fundamentally reshape banking operations—shifting from department-centric management structures to new organizational forms centered on professional expertise and customer value," Gong noted.

Looking ahead, DBS Group will focus more on developing employees' augmented capabilities and nurturing specialized knowledge workers. Over the next five to ten years, banks that align their organizational structures with AI adoption will gain greater competitive advantages in seizing market opportunities, Gong emphasized.

To realize this vision, he outlined a three-step approach: setting transformational goals, designing operational models and roadmaps, and implementing changes incrementally.

Gong also revealed that DBS Group's future call center workflows will be driven by an "intelligent agent hub" for automated classification and routing. Human roles will evolve into three key functions: 1) A "gatekeeper" team of senior experts to validate AI outputs; 2) A "governance team" for model monitoring and compliance; 3) An "enablement team" to integrate new knowledge into systems and verify outcomes.

Critical steps for this operational shift include data governance, isolating key datasets, system enhancements, and integrating AI into daily operations with support from expert data scientists.

"Gen-AI transformation is a multi-year journey—likely requiring three to five years or longer—but we've begun taking concrete steps," Gong concluded.