Tech giant Huawei is launching its agentic banking framework to assist financial institutions in achieving true hyper-personalisation and accelerating legacy application modernisation.
Speaking at The Huawei Intelligent Finance Summit in Shanghai, Jason Cao, CEO of the Huawei Digital Finance Business Unit, outlined how autonomous AI agents will restructure traditional front, middle, and back-end banking architectures into an AI-native ecosystem.
The shift toward agentic banking marks a transition where AI evolves from a basic productivity tool into an active, autonomous colleague. This technological leap enables financial institutions to dismantle long-standing operational limitations and scale high-value services across their entire user base.
“In the past, banks relied heavily on the 80/20 rule, where 20 per cent of customers created 80 per cent of the profit,” Cao explained during an interview with Richie Santosdiaz, Economic Development Correspondent at The Fintech Times. “When agentic AI arrives with its strong capabilities, everyone could be treated as a VIP customer, meaning hyper-personalisation can finally come true”.
To support this autonomous capability, Huawei presented a redefined three-layer technology stack designed to replace rigid legacy systems. The first layer focuses on customer experience, transitioning user interactions from a Graphical User Interface (GUI) to a Language User Interface (LUI). This layer relies heavily on intent recognition and long-term memory. Unlike current systems that merely record transaction histories, an agentic system retains the context of historical customer interactions over several years, allowing the AI to comprehend exact user needs instantly.
The second layer manages multi-agent collaboration. This layer shifts the operational model from human workers utilising separate tools to an integrated environment where humans work alongside AI colleagues. These specialised AI agents automatically collaborate to execute complex workflows, removing friction from standard banking processes.
The third layer redefines the decision model, moving financial institutions away from basic rules and structured data toward ontology and knowledge-based decision-making. This layer allows banks to utilise massive amounts of unstructured data, such as documents and images, while transforming the decades of experience held by senior risk management experts into reproducible digital knowledge models.
Implementing agentic banking requires financial institutions to navigate the coexistence of legacy platforms, cloud-native applications, and new AI-native systems. Cao noted that while cloud-native applications serve as a vital bridge, the primary obstacle remains modernising monolithic core banking systems that are often decades old. To resolve this, Huawei focuses on accelerating application modernisation by utilising AI to automate the conversion of legacy code, such as translating COBOL to Java.
Huawei is bringing its extensive engineering experience from mainland China, where massive tier-one banks successfully migrated hundreds of millions of users to modern architectures with zero service interruption, to global markets. International financial institutions frequently require multi-day service shutdowns to complete core cutovers, but Huawei leverages synchronized data architectures to shut down legacy infrastructure seamlessly. The company has already commenced these modernisation journeys with leading banks across Southeast Asia such as Singapore and Thailand, Africa, Latin America, and the Middle East.
Tactically, Huawei advises financial institutions to remain pragmatic and flexible by focusing on specific, high-value commercial business cases rather than attempting immediate, large-scale overhauls. For instance, a bank in Thailand experiencing 40,000 daily fraud reports utilised AI agents to handle case loads exponentially faster, allowing a team of 200 bankers to resolve complex transactional relationships that humans could not easily identify manually. In the Middle East, financial institutions are deploying AI agents to automate document review processes, significantly improving operational efficiency and customer satisfaction within months of implementation.
Ecosystem collaboration forms a major component of this global strategy through the RONGHAI Global Partnership Program, an initiative combining finance and inclusion. The programme encompasses 180 solution partners across 30 distinct categories, creating a four-win model that benefits financial institutions, independent software vendors (ISVs), local system integrators, and Huawei. Cao said, “Nobody alone can do this job. The banks cannot and we cannot. We have to work together and hence why we at Huawei have the ‘RONGHAI’ plan.”
To support the long-term scalability of these AI platforms, Cao recommended that banks adopt a hybrid AI architecture. This model keeps sensitive customer data and core business logic on-premises to satisfy strict regulatory compliance and security demands, while utilising public cloud infrastructure for less sensitive tasks. To address the global shortage of technical expertise, Huawei also announced a talent development initiative aiming to train 10,000 AI professionals over the next three years, ensuring financial institutions possess the necessary skills to sustain their agentic banking ecosystems.
Beyond just China, Huawei, even beyond from being a global player in consumer products, is also showing its know-how and talent across the financial services sectors as well.
The post Modernise Core Banking Systems to Deliver Hyper-Personalisation Through Agentic AI appeared first on The Fintech Times.
