Large data applications need to connect the industry and the financial integration Machine essay

How hot is the big data? From the Spring Festival Big Data, the Spring Festival gala Big Data to "two sessions" on the big Data national strategy heated discussion, so, big data from a technical market hot word, suddenly "like the night Breeze come, fly into ordinary people home".

In the technology market, discussions about whether big data is "new bottles of old wine" or "disruptive innovation engines" never stop. In fact, when the value of information contained in the data lifecycle is fully released, no one cares whether the technology behind it is "big data" or something else. What people experience in real life are all kinds of "big data fruits" that can be seen or invisible, such as various related recommendations on shopping websites, display of goods in shopping malls, personalized derivative services based on data, and even "where is the best beer and fried chicken?" "This trend.

The essence of Internet finance-Big Data

2014 Spring Festival, "micro-letter Red envelopes" and pay treasure "let red envelopes Fly" for a while, its thunder must let people forget that year crazy indiscriminate "gifts only receive Brain Platinum", successfully occupy a variety of information channels headlines. Together with the "Internet finance", which is once again pushed to the forefront of public opinion, the service innovation of traditional financial industry has once again become the hot topic of all. After all, to the micro-letter red envelopes, for example, it handed over the answer: from New Year's Eve to lunar 16 o'clock, participated in the robbery of red envelopes 5 million users, a total of more than 75 million packets of red envelopes, the total number of more than 20 million, the average pick up to 94,000 per minute ...

Trading data, credit data, social data, mobile data and so on, behind these data assets, is to imagine a borderless, huge "information treasure." Who can better master and analyze these data, who will be able to better mining the data behind the business value, realize points into "gold." In this respect, Internet companies are "Pathfinder" is also "Pioneer", the traditional industry products, service innovation will follow, after all, the big data is behind the hands.

As China's earliest completion of information construction industry, the financial industry has accumulated considerable data over the years. In recent years, the financial industry's own customer data, transaction data, risk management data and data from the network are exploding, large data application prospects are very broad.

However, how to draw on the success of internet finance, the use of large data technology to innovate financial business and mode, is a lot of financial institutions have been thinking about. The traditional financial industry has a long business chain and many business links, which brings the present situation of large amount of data, dispersion and heterogeneity. On the one hand, through the adoption of virtualization, distributed computing, storage technology, across the organizational structure, business system limitations, with large data to achieve business transformation and service innovation is imperative; On the other hand, there are many technical schools in the large data technology market, which bring the technical compatibility problems of hardware, software and so on, resulting in the deployment of large data, Business development and operation and other links, the user faces a long construction cycle, debugging, two times the difficulty of high development problems.

Insufficient internal force of large bank data

Hequan said: "Virtually all banks themselves are big data companies." "According to statistics, the average bank data storage is the highest in all industries, the domestic bank structured data average annual growth of about 100TB, close to one-twentieth of foreign countries, unstructured data about 1PB." Among them, ICBC's enterprise-class data Warehouse storage more than 350TB, the accumulation of data is more than 4.5PB, BOC, CCB, UnionPay, Bank, Huaxia, postal and so on also have a large amount of data.

What can banks do with these vast amounts of data? Based on big data, McKinsey reckons, banks can increase their clients ' share of loans by one-fold and reduce loan losses by One-fourth. At present, large data in the banking sector has a wide range of risk management applications, such as the delivery of the Pacific Credit Card Center adopted the automatic credit approval decision engine, the average balance of quality accounts and interest income increased by 20% and 10%, approval efficiency increased by 30%; After using large data for risk management, ICBC Manual supervision of the whole line of daily work volume from 8.598 million to 54,000, the entire line released more than 5,900 business supervisors.

Large data can also effectively reduce the bank's customer churn rate. Merchants Bank through data analysis and identification the high value of credit card customers often appear in Starbucks, DQ, McDonald's and other places, through the "Multiple integral cumulative", "Points store Exchange" and other activities to attract quality customers; Through the construction of customer churn early warning model, the loss rate of 20% of the top customers to sell High-yield financial products to retain, So that the gold and gold Sunflower Card customer turnover rate decreased by 15 and 7% respectively.

However, the ability of large data collection and analysis in the banking industry still needs to be strengthened, especially in the aspects of voice analysis, video analysis, flow analysis and position analysis, there is still a big gap between the average level of banks and other industries.

From the traditional structural data analysis and application, to the industry's internal integration of multiple structure type of large data processing, and then to the future cross-sectoral, Cross-industry integration Analysis, "We are in the industry of heterogeneous data sharing and value release phase, and in the future for quite a long time will be at this stage." "At this stage, there will be a lot of new technologies and new products, whether it's hardware or software," says Wang Endong, senior vice president of Wave group. This is a rare ' historical new opportunity ' for system vendors and industry ISVs in the downstream industry chain. ”

When the general technology is not enough to support the business process reengineering and value release, combining the characteristics of industry requirements, technology and product integration and optimization is a more realistic and feasible approach. A few days ago, wave group and the domestic leading financial information solutions suppliers South-day information jointly launched a wave of clouds of cloud finance large data machine, this is the first domestic market for the financial industry large data applications integrated products. It is understood that the system has been in a state-owned commercial banks to complete the historical inquiry business, test results outstanding performance.

Large data industry applications to the "atmosphere"

In view of the large data market positioning and analysis, the wave strategy can be described as "both hands grasp, both hands must be hard." On the one hand, the tide in the basic software and hardware platform to continue to increase investment in research and development, and some key technologies to the integration of products. On the other hand, the continuous establishment and deepening of cooperation with the domestic large-scale ISVs, improve the industrial layout, has become a big wave of data strategy important strategy. An open and flexible foundation platform combined with industry applications to form an integrated solution can effectively help users cross the threshold of large data applications and create new business value.

It is understood that, based on the distributed open architecture of the Wave Cloud Valley system software, clouds of financial large data integration machine using the share nothing architecture design, with flexible expansion, that is, the performance of strong, safe and reliable four characteristics, its application of the FPGA acceleration technology, SSD accelerated caching technology, such as a number of optimization technology to achieve a large increase in overall performance, data import speed can reach millions per second, random query speed is as high as Bai per second, than the industry average performance level increased by 20%. and share nothing architecture design, so that the system can be flexible expansion to 4000 nodes, to achieve the top Internet companies large data technology implementation capabilities.

At the same time, the sea of clouds financial data one machine application many based on the Tide shuttle K1 Host Security technology, has self-developed more than 50 invention patents and 2 software copyrights, RAS features more than 20, and modular security system design and operating system security enhancement system, To ensure that the product conforms to the national information security level to protect the third level of certification standards.

Large data one machine, one machine system software, large data industry customization machine, as the first launch of large-scale data platform solutions for domestic manufacturers, the wave of large data strategy clear, the product is enough "grounding gas." When the basic platform and industry application fully grafted, data flow and value change will be from the real sense of the "paper" stage, large data applications to the industry depth, "data-centric, the promotion of industry applications from data to service transformation" dream will be taken into reality.