Modern Core Banking with Agentic AI Promises New Era of Financial Services

Huawei is launching an agentic banking framework aimed at helping financial institutions achieve true hyper-personalisation and accelerate modernisation efforts. According to Jason Cao, CEO of Huawei Digital Finance Business Unit, this approach will restructure traditional bank architectures into AI-native ecosystems where AI functions as autonomous colleagues.

“In the past, banks relied heavily on the 80/20 rule,” Cao explained in an interview with The Fintech Times. “When agentic AI arrives with its strong capabilities, everyone could be treated as a VIP customer.” This shift promises to move beyond basic productivity tools toward active AI agents that can handle complex tasks and scale high-value services across entire user bases.

Three-Layer Technology Stack

Huawei’s solution features a redesigned technology stack:

  1. Customer Experience Layer: Transitioning from graphical interfaces to language-based interactions with advanced intent recognition and long-term memory - allowing AI to understand customer needs based on years of interaction data.
  2. Multi-Agent Collaboration Layer: Integrating human workers with specialised AI agents that automatically collaborate on complex workflows, removing friction from standard banking processes.
  3. Decision Model Redefinition: Moving beyond structured data toward knowledge-based decision-making that can utilise unstructured data like documents and images, while encoding expert institutional knowledge into digital models.

Addressing Modernisation Challenges

The biggest obstacle remains modernising legacy core banking systems - often decades old - but Huawei is using AI to automate code conversion (e.g., COBOL to Java) and leverage synchronized data architectures for seamless transitions with zero service interruption.

The company has already begun these journeys in Southeast Asia, Africa, Latin America, and the Middle East, advising institutions to focus on specific high-value use cases rather than attempting large-scale overhauls all at once.

For example, a Thai bank experiencing 40,000 daily fraud reports used AI agents to handle caseloads exponentially faster, while financial institutions in the Middle East are automating document review processes with significant efficiency gains.