2021 Alibaba cloud financial data Intelligence Summit — experience of Alibaba data middle office construction

Time:2022-6-20

Introduction:Alibaba data middle office’s solution is to achieve intelligent data approval on the basis of introducing data security level marking, and make intelligent processes replace cumbersome manual work through trusted model construction and risk quantification. Taking dataphin as an example, as the product output of Alibaba Group’s data governance methodology based on internal practice, it can provide enterprises with the ability of one-stop data acquisition, construction, management and full life cycle management, so as to help enterprises significantly improve the level of data governance and build an enterprise level data platform with reliable quality, convenient consumption and safe and economic production.

-For more information about digital intelligence transformation and data center, please joinAlibaba cloud data China Taiwan exchange group – Digital Intelligence ClubAnd follow the official wechat official account (scan QR code orClick here to join

-Alibaba cloud data platform official websitehttps://dp.alibaba.com/index


This year marks the end of the three-year development plan, and the importance of digital transformation of financial institutions, including banks, securities and insurance, is further highlighted.

At the 2021 Alibaba cloud financial data Intelligence Summit held today, Wang Sai, head of Alibaba cloud data center, revealed that Alibaba’s experience in data center construction may bring some lessons to financial enterprises’ digital intelligence transformation.

2021 Alibaba cloud financial data Intelligence Summit -- experience of Alibaba data middle office construction

Wang Sai, head of Alibaba cloud data center

Six experiences of Alibaba data platform construction

In 2015, Alibaba officially put forward the data middle office strategy, which is also the first time that the concept of data middle office has appeared in China – but Alibaba actually did something earlier.

Wang Sai said that as early as 2011, Alibaba had reorganized and rebuilt its internal data team to build a data center based on “business segment + analysis dimension”. In 2013, oneservice, a unified data service middleware, was officially born. After in-depth processing, oneservice can provide unified data services for front-end businesses.

In his opinion, Alibaba’s data platform construction is by no means a cluster, but a digital intelligence transformation path that must be completed in the face of complex scenarios and diverse needs within Alibaba group.

These scenarios and requirements can be summarized into six categories: data quality and safety, data value, product tool precipitation, cost control, organization and operation, and quality and assessment.

Data quality and security mainly focus on four aspects. The first is consistency. In the face of the same data, the definition caliber of each business is inconsistent, which brings great trouble to the later development, analysis and application. Therefore, the first thing the data center should solve is the standardized definition of indicators, and on this basis, it can realize the overall construction of code and the output of data results; Secondly, there is the depth of data assets. Through the deep integration and communication of data, the data center can provide all-round market information for front-line employees and conduct value evaluation; Thirdly, the timeliness of data is guaranteed. Through data operation and maintenance baseline management and mobile office collaboration, the data middle office can ensure that the business department can obtain multidimensional process and result data in a timely manner. Even in the mobile office scenario, it can also ensure that relevant data can be viewed in real time; Finally, the focus is on data circulation security. The core is to solve two problems: the definition of data security approval authority and the reduction of approval workload on the basis of ensuring data security. Alibaba data center’s solution is to realize data intelligent approval on the basis of introducing data security grade marking. Through the construction of trusted models and risk quantification, intelligent processes can replace cumbersome manual work.

On the other hand, data value is mainly reflected in helping platform growth, business growth and employee efficiency through data empowerment.

It is worth noting that during the construction of Alibaba data platform, a very rich product matrix has been derived to deal with complex business scenarios and personalized job needs, such as the media screen for the “double 11” scenario, analytical data products for management decisions, etc.

At the same time, actively build a data talent training system, establish a “Data Committee” so that employees in various business data posts can form efficient linkage and trust, and condense and spread a scientific and effective data culture.

In the whole process of building the data middle office, Wang Sai stressed, “we have also refined the explicit expression of the data middle office value, and repeatedly verified the setting of KPIs, personnel and budgets from the multi-dimensional perspectives of user value and experience, asset precipitation and operation. This is a dynamic process.”

The ability to comprehensively output data through Alibaba cloud

After years of internal practice, Alibaba data center officially passed Alibaba cloud’s ability to fully open its data center in 2018. So far, it has been successfully implemented in finance, retail, government affairs, the Internet and other industries.

Liuweiguang, general manager of Alibaba cloud’s new finance & Internet business unit, said in a media interview earlier that Alibaba cloud data has two unique advantages for enterprises, “The first is the richness of tools. Alibaba cloud data mid tier products integrate all the tools on the market, from the tool level, from the data processing level, to the upper application level, to the data use level, to the bi level, to the decision-making level, which can be said to cover everything.”

2021 Alibaba cloud financial data Intelligence Summit -- experience of Alibaba data middle office construction

Liuweiguang, general manager of Alibaba cloud new finance & Internet business department

At present, Alibaba cloud data center has formed a core product matrix dominated by dataphin, quick Bi, quick audience and other products.

Taking dataphin as an example, as the product output of Alibaba Group’s data governance methodology based on internal practice, it can provide enterprises with the ability of one-stop data acquisition, construction, management and full life cycle management, so as to help enterprises significantly improve the level of data governance and build an enterprise level data platform with reliable quality, convenient consumption and safe and economic production.

At the same time, dataphin provides a variety of computing engine support and expandable open capabilities, which can adapt to the platform technical architecture and personalized demands of all walks of life.

Focus on the financial industry. As an industry with an early start in digitalization, the construction cycle and history of data platforms in banking, securities, insurance and other industries are no shorter than those in the Internet industry. They have accumulated some experience in data use. However, it is still difficult to avoid the pain of digital intelligence transformation.

The core is reflected in four aspects: data standards, data quality, corresponding demand and cost resources:

  Data standard issues: chimney development and local business service support lead to frequent problems with the same name and different dimensions of indicators; Historically, different business systems have been launched iteratively, and the attribute codes of the same object are inconsistent;

  Data quality problems: repeated construction leads to long task chain, numerous tasks, tight computing resources and poor data timeliness; There is a gap between the precipitation of documents defined by caliber sorting and the implementation of development code, and the risk of ensuring data accuracy is high;

  Demand response problem: the chimney development cycle is long, the efficiency is low, and the application-oriented service is insufficient, resulting in slow business response, dissatisfaction with the business, and no precipitation and growth of technology; There are not enough talents who understand both business and data. They need to understand that the development and implementation involve a lot of communication, and the service efficiency is poor;

  Cost resource issues: the repeated construction of chimney development wastes technical resources; It is more difficult to go online and offline. The source system or business changes cannot be reflected in the data in a timely manner. In addition, the data is not standard, which makes it more difficult to develop and maintain. At the same time, a large number of useless calculations and storage cause a waste of resources.

2021 Alibaba cloud financial data Intelligence Summit -- experience of Alibaba data middle office construction

This also coincides with the difficulties faced by Alibaba in the construction of data platform.

The data governance methodology put forward by Alibaba through practical experience can help enterprises clarify the management idea of the whole data life cycle, and implant it into the product dataphin (intelligent data construction and management) to provide services for enterprises through Alibaba cloud.

For this reason, in addition to the data integration, development, release, scheduling and operation and maintenance capabilities involved in the whole link of big data processing, dataphin will also provide enterprises with data specification definition, logical model definition, code automatic generation and data subject service capabilities, and effectively complete the construction of data.

Based on the ability of dataphin products, CAITONG securities has opened up multiple existing system data, realized timely data access and unified standards, and completed more than 300 data labels including “financial attribute” and “product type” based on the market form after integrated processing.

After 7 months of CO creation and co construction, digital Wo technology has completed the reconstruction and upgrading of the traditional data warehouse system, unified the data asset management platform, comprehensively combed the company’s core business processes with the help of the project, connected all business processes and the corresponding data behind them, and unified a set of information systems. Each business process can see the specific data warehouse table, indicators, current values of indicators, month on month, year-on-year and other information on the information platform, Once these indicators are abnormal, they can quickly automate attribution, locate problem links, and creatively establish a data operation model.

Wan Peng, head of big data of digital Wo technology, said: “the Alibaba cloud data console comes with the one service data interface service. Previously, the data platform we developed was too long and complex to provide data interfaces for online businesses. At present, the data input and output links built through the Alibaba cloud data console product dataphin are smooth and fast, and the product comes with its own integration channel.”

In addition to dataphin, Alibaba cloud data center also provides professional products and services for financial institutions in many fields, such as global operations and data visualization analysis.

2021 Alibaba cloud financial data Intelligence Summit -- experience of Alibaba data middle office construction

The essence of big data is data fusion. Alibaba cloud data center associates and fuses their isolated data, and constructs a data asset category system through abstraction and processing, so as to endow data with deeper semantics and value and insight into the nature of things.

In the past year, it has completed a major upgrade in technical capability – Taking the lead in realizing “Lake Warehouse Integration” and leading the evolution direction of the next generation big data processing platform through collaborative work architecture; By integrating graph calculation, time series calculation, privacy calculation, etc., the platform’s data intelligence capability is greatly improved to help customers make intelligent decisions.

In the year when the three-year development plan is coming to an end, the data center will certainly give more possibilities to the digital intelligence transformation of the financial industry.


Data middle office is the only way for enterprises to intellectualize. Alibaba believes that data middle office is an intelligent big data system integrating methodology, tools and organization, which is “fast”, “accurate”, “complete”, “unified” and “accessible”.

Currently, Alibaba cloud is exporting a series of solutions, includingCommon data midrange solutionRetail data midrange solutionFinancial data middle office solutionInternet data midrange solutionGovernment data middle office solutionAnd so on.

The Alibaba cloud data midrange product matrix takes dataphin as the base and quick series as the business scenario, including:

Official site:

Official website of data centerhttps://dp.alibaba.com

Nailing communication group and wechat official account

Copyright notice:The content of this article is spontaneously contributed by Alibaba cloud real name registered users. The copyright belongs to the original author. The Alibaba cloud developer community does not own the copyright, nor does it assume corresponding legal responsibilities. Please refer to Alibaba cloud developer community user service agreement and Alibaba cloud developer community intellectual property protection guidelines for specific rules. If you find any content suspected of plagiarism in the community, fill in the infringement complaint form to report. Once verified, the community will immediately delete the content suspected of infringement.