Huawei has launched its Fintelligent AI Solution globally for financial institutions, combining agent development, AI token operations and data-knowledge infrastructure as banks move from isolated AI applications toward enterprise-wide agentic banking capabilities. The solution was unveiled at HUAWEI CONNECT 2026 during Huawei’s Global Finance Summit.

Huawei is expanding its financial-services AI portfolio with the global launch of the Fintelligent AI Solution, a platform architecture designed to help financial institutions build, operate and scale enterprise AI systems for what the company calls “Agentic Finance.”

Unveiled during HUAWEI CONNECT 2026 in Shanghai, the solution is built around three components: Agent Factory, Token Factory and Data-Knowledge Factory. Huawei says the architecture is intended to address three major challenges associated with large-scale financial AI deployment: security, scalability and sustainability.

The launch reflects a shift in how financial institutions are approaching AI. Rather than deploying individual generative AI applications, banks are increasingly looking at the underlying infrastructure needed to manage multiple AI agents, models, data sources and workloads across the enterprise.

Jason Cao, CEO of Huawei’s Digital Finance Business Unit, framed that shift as a move from competing primarily on model capabilities toward building enterprise-wide AI system capabilities. Huawei’s proposition, “Own Your AI, Own Your Intelligence,” emphasizes control over the architecture, operation and evolution of AI systems.

The Agent Factory forms the agent-management layer of the platform. It is designed to help financial institutions create, deploy, operate and govern AI agents at scale while enabling organizations to accumulate and reuse agent capabilities.

A central component is openJiuwen, Huawei’s open-source agent platform for multi-agent collaboration. Huawei describes it as having a financial-grade production kernel with capabilities for long-running operations, financial components, and integrated development and operations. The company says the platform has been deployed at more than 10 financial institutions, with more than 30 partners participating in its development. Huawei also claims that a single customer environment can support more than 1,000 agents and handle tens of thousands of concurrent connections.

That architecture puts agent governance alongside agent creation. For banks, this distinction is important because deploying hundreds or thousands of autonomous systems can create challenges around permissions, data access, monitoring and accountability. Huawei has separately emphasized that enterprise agents need to be organized around business processes rather than deployed as disconnected applications.

The Token Factory addresses another emerging infrastructure issue: the cost and operational complexity associated with AI inference. Huawei launched TokeNexus, which manages token planning, production, scheduling and optimization across the AI lifecycle.

The company positions the system as a way for financial institutions to balance AI performance with compute costs and supply. For cloud deployments, Huawei offers Model-as-a-Service, while its on-premises infrastructure includes products such as Atlas 850E, TaiShan 950 and OceanStor A800 for inference, agent execution, memory and retrieval workloads.

The third component, the Data-Knowledge Factory, focuses on turning financial data, documents and business rules into knowledge that AI agents can use. Huawei also introduced its Financial Agentic Data Solution, built around data and knowledge pipelines that continuously supply agents with information.

This represents a change from traditional enterprise data architectures, where information is primarily structured for employees and business applications. In an agentic environment, data must also be prepared for machine-driven reasoning and task execution.

Huawei is also applying the approach to legacy modernization. Its AI Coding Solution uses an ontology-centric approach to extract business rules embedded in legacy application code and convert them into reusable knowledge assets for AI systems.

The broader strategy is consistent with Huawei’s earlier work on financial AI infrastructure. At its Intelligent Finance Summit in May, the company described agentic banking as a progression requiring AI infrastructure, applications, data intelligence and resilient connectivity rather than isolated AI deployments.

For financial institutions, the appeal of this architecture is its attempt to connect several layers of enterprise AI management. Agent capabilities, compute resources, data and institutional knowledge are treated as parts of one operating environment rather than separate technology projects.

There are also practical constraints. Banks deploying agentic systems need to address model accuracy, cybersecurity, regulatory compliance, data governance, access controls and operational resilience. Scaling the number of agents does not automatically translate into business value, particularly when agents interact with sensitive financial data or core banking systems.

Huawei’s strategy therefore goes beyond generative AI interfaces. Its Fintelligent AI Solution is positioned as infrastructure for building AI capabilities that can be operated continuously and expanded across banking functions.

The company says it has served more than 7,100 financial customers across more than 80 countries and regions, including 54 of the world’s 100 largest banks. Those figures are Huawei’s own customer claims rather than independently verified market-share data.

As financial institutions move toward agentic applications, the competitive focus is likely to extend beyond model selection to the infrastructure surrounding those models. Platforms that can combine agent development, enterprise knowledge, AI compute and governance could become an increasingly important part of banking technology stacks.

Huawei’s Fintelligent AI launch places the company directly in that infrastructure conversation, with its proposition centered on giving financial institutions greater control over how enterprise AI is built, operated and evolved.

Market Landscape

Financial AI is moving from isolated use cases toward enterprise platforms capable of supporting multiple agents, models and business processes. Huawei’s strategy follows this direction by combining agent infrastructure with AI compute and data-knowledge management.

The company has also been developing adjacent infrastructure for AI-heavy financial workloads. Its Xinghe Intelligent Financial Network, upgraded earlier in 2026, was designed around the increasing need to generate, transmit and consume AI tokens efficiently across financial workloads.

The competitive landscape includes hyperscalers, enterprise software companies, AI infrastructure providers and financial-technology vendors. The key technology questions are increasingly shifting toward governance, interoperability, inference economics, data readiness and the ability to operate AI reliably at enterprise scale.

Top Insights

• Huawei’s Fintelligent AI Solution combines agent development, AI token operations and data-knowledge infrastructure for financial institutions.
• openJiuwen provides the open-source agent foundation, while TokeNexus targets lifecycle management of AI token production and consumption.
• The Data-Knowledge Factory is designed to convert financial data, documents and business rules into reusable intelligence for AI agents.
• Huawei is positioning enterprise AI infrastructure, rather than individual models, as the foundation for large-scale agentic banking.
• Security, governance, compute economics and data quality remain critical challenges as banks expand AI from pilots into production.

Get in touch with our fintech expert