数据沿袭工具改善风险管理，推动合规

瑞银

瑞银是一家总部位于瑞士的跨国投资银行和金融服务公司。它成立于 1862 年，是著名的瑞士银行业最大的参与者。瑞银在 50 多个国家设有办事处，拥有超过 66,000 名员工，2018 年总资产超过 9,580 亿美元。瑞银在五个主要业务领域开展业务：个人银行业务、财富管理、企业和机构客户、投资银行业务和资产管理。

在 2007 年全球金融危机之后，瑞银需要遵守为加强风险数据汇总和内部风险报告系统而制定的法规。具体而言，瑞银寻求遵守巴塞尔银行监管委员会发布的标准 239 (BCBS 239)。

UBS 构建了一个名为 Group Data Dictionary (GDD) 的应用程序作为其数据沿袭和数据治理工具。

第一次迭代是在 Oracle 上构建的，但 UBS 很快发现了 RDBMS 方法的局限性，该方法依赖 JOINS 跨表连接数据。 UBS 决定它需要一个更好的解决方案，以适合创建实时数据沿袭可视化和导出沿袭信息以通过 Excel 进行临时分析。

UBS 将 Neo4j 与 Oracle 系统同步，从初始数据加载开始，然后执行增量同步，从 Oracle 表中读取事务并实时写入 Neo4j。

UBS 使用 Neo4j 评估数据沿袭并在 GraphJSON 中描述结果。此信息流入 D3.js 可视化工具，以将数据呈现为沿袭图。

拥有所有元数据有助于轻松报告。当出现特定问题时，这些数据可用于临时报告，并且可以将整个谱系导出到 Excel。

• The implementation of Neo4j has significantly improved UBS's risk management and compliance processes. The company can now trace the lineage of all metrics for all initiatives across the bank in real time, a task that would have been impossible using traditional technologies. The solution has eliminated the complexities involved with traversing multiple tables in a relational database, with results now easily obtained and displayed in an intuitive graph format. Furthermore, Neo4j's flexibility in not relying heavily on predefined schema or indexes allows it to keep pace with constantly changing business analysis requirements. The tool has enhanced UBS's understanding of the flow of data through the organization, the impact of changes in one application on the entire organization, and how errors can propagate through the system.

• UBS was able to meet regulatory requirements within a short window and with limited resources.
• The solution provides real-time response time and agility, with lineage spanning dozens of levels of entities and dependencies.
• The data lineage tool is always up-to-date with the enterprise system thanks to synchronization.