In the FinTech era — characterized by the explosion of data, both structured and unstructured — Huawei works with ecosystem partners to provide end-to-end data plane solutions tailored for financial customers.
Huawei Converged Financial Data Lake integrates products from multiple vendors and provides several differentiated advantages. It effectively helps banks overhaul and reconstruct their capabilities, to offer precise customer acquisition tools and real-time risk controls. The Solution also helps banks build leaner operations in front, middle, and back offices, and allows them to intelligently craft personalized products and experiences for their own customers.
Consisting of an innovative application layer, a converged platform layer, and an intelligent infrastructure layer, the solution offers three types of service applications: for marketing, for operations, and for risk control.
For marketing, applications include customer profiles, product profiles, personalized recommendations, precision marketing, and intelligent information pushes.
At the operations level, applications include intelligent warning of key indicators, after-sales feedback analysis, and intelligent data center Operations and Maintenance (O&M).
Finally, risk control applications offer real-time online credit approval, real-time transaction fraud prevention, and customer default prediction.
Full data, service, and architecture convergence supporting: the analysis of structured and non-structured data; compatibility with traditional and innovative services; unified O&M management of multiple clusters, interconnection between MPPDB and Hadoop, computing pushdown, collaborative analysis, and compatibility with x86/ARM servers.
Open architecture and platforms decoupling software from hardware. Clusters are provisioned as services on both private and public cloud, supporting cloud-based evolution. Logical clusters load various services. With Artificial Intelligence (AI), the solution enables GPU acceleration and is compatible with many AI platforms.
Implementation of full parallel computing (nodes, operators, and instruction sets), improving performance by up to 10 times. Online capacity expansion, without service interruption, achieved within an hour. Support for standard ANSI SQL, with quick, smooth migration tools for Teradata and Oracle programs.