Customer Service Automation · Document Processing · Marketing Personalization · Risk Management
Rapid growth in digital banking, mobile payments, online wealth services and inclusive finance increased transaction and compute demand. Changsha Bank also needed to operate heterogeneous AI compute efficiently, manage models of different sizes, and scale AI into financial workflows while modernizing its core systems and cloud-native architecture.
Reported impact
The bank reports AI support for more than 40 financial scenarios, threefold compute utilization, development cycles shortened from months to weeks, twice the inference performance, improved wealth-service efficiency and customer-profile accuracy, and greater than 80% accuracy for the reported credit debit-deduction task.
Changsha Bank worked with Huawei on a cloud-native, distributed technology foundation for its banking workloads. Huawei Cloud Stack provides a full-stack, multi-cloud platform designed for mixed processor architectures; Kunpeng servers host the migrated online-banking and network-connect core systems. For AI, the bank combines Huawei's hybrid-cloud platform with an Ascend AI computing base and a five-part system for compute, models, knowledge, R&D and evaluation. This addresses heterogeneous compute scheduling and management of small and large models, supporting AI across more than 40 financial scenarios. Reported deployments include credit, risk-control and wealth-marketing use cases, plus the Le series AI assistants for wealth questions, recommendations and planning, and intelligent credit-report generation. The page reports compute-utilization and inference gains but does not detail data ingestion or storage design.
Compute utilization increased to three times and the R&D cycle shortened from months to weeks. (original: 算力利用率提升至三倍，研发周期从月缩短至周。)
Inference performance increased to twice the previous level. (original: 凭借昇腾推理加速技术实现推理性能提升至两倍。)
The system supported more than 40 financial scenarios and average monthly calls exceeded 100,000. (original: 长沙银行已成功孵化信贷、风控、金融等40余个金融场景，月均调用超过十万次。)
Wealth-service efficiency improved by 50% and customer-profile accuracy reached 85%. (original: 服务效率提升50%，客户画像精准度达到85%。)
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Huawei Cloud Sample Archived · Apr 20, 2026 Postal Savings Bank of China Postal Savings Bank of China needed a scalable AI infrastructure and unified model platform so it could automate customer service, trading, document review, and risk workflows without creating disconnected AI pilots… RFQ volume exceeded 1.5 trillion yuan AI computing and data center · Huawei Cloud Stack · Youzhi large model 
Google Cloud Sample Archived · Jan 16, 2026 Hong Leong Bank Hong Leong Bank needed to replace slow manual banking operations and limited chatbot performance with a secure AI ecosystem that could support dynamic customer support, document-heavy trade operations, sanction… 30% efficiency gain Agent Development Kit (ADK) · Cloud SQL · Compute Engine 
In-house Sample May 8, 2024 Mapfre Mapfre wanted to improve customer experience and streamline insurance operations such as claims handling, risk assessment, underwriting support, document management, internal knowledge access, and digital content… 75 use cases under study for the application of generative AI ControlExpert · Cyberwrite · Holistic AI 