The digital transformation of securities firms has entered the fast lane, and multiple policy documents have been released one after another, requiring the improvement of financial technology governance levels, increased investment in technology funds, and deepening digital transformation.
At the same time, affected by the downward macroeconomic environment, the homogeneous competition among securities companies has intensified, and the growth of traditional brokerage business has been under pressure. Securities companies urgently need to find new business growth points.
Under the dual influence of policy drivers and performance pressure, meeting customer needs through digital means, promoting business innovation, improving operational efficiency, and ultimately enhancing the market competitiveness of securities companies are the core goals of the digital transformation of securities companies at this stage.
Against this background, in order to help securities companies explore the actual implementation experience of digital and intelligent transformation, and discover high-quality manufacturers, on September 8, 2023, AiAnalysis officially released the "2023 AiAnalysis·Securities Digitalization Practice Report".
Click "Read the original text" in the lower left corner of the article to download the full version of "2023 Love Analysis･Securities Digitalization Practice Report".
Report key findings
Brokerages are facing the challenge of wealth management transformation, and digitalization is an important path to achieve it. Brokerages should make full use of artificial intelligence technology, especially large models, to empower customer acquisition, investment strategy construction, investment portfolio implementation, and investment education in the wealth management value chain.
In order to achieve cross-department collaboration and efficient operations, the business middle office is the key for securities firms to achieve IT reshaping. In view of the wide range of upper-level application scenarios supported by the business platform, securities companies can adopt a strategy of starting from easy to difficult, starting from a single scenario and then gradually covering all scenarios, and select document writing scenarios that cover a wide range and do not affect business continuity.
Intelligent IT operation and maintenance for securities companies is the future trend. In this regard, during the supplier selection process, securities firms should select vendors with data management capabilities and intelligent computing engines to meet intelligent needs.
Report typical cases
Relying on intelligent core technology, Guojin Securities uses Yuxin Intelligent Writing Center to improve multiple business capabilities
The intelligent operation and maintenance analysis platform accelerates the evolution of Essence Securities’ intelligent operation and maintenance analysis capabilities< /span>
01 Report Summary
The digital transformation of securities companies has entered the fast lane. The "Financial Technology Development Plan 2022-2025" and the "Securities Company Network and Information Security Three-Year Improvement Plan (2023- The successive release of policy documents such as "2025)" requires improving the level of financial technology governance, increasing investment in technology funds, and deepening digital transformation, providing clear guidance and direction for the digital transformation of securities companies.
At the same time, affected by the downward macroeconomic environment, the homogeneous competition among securities companies has intensified, and the growth of traditional brokerage business has been under pressure. Securities companies urgently need to find new business growth points.
Under the dual influence of policy drivers and performance pressure, meeting customer needs through digital means, promoting business innovation, improving operational efficiency, and ultimately enhancing the market competitiveness of securities companies are the core goals of the digital transformation of securities companies at this stage.
Guided by business, the current digital transformation of securities firms presents the following three major trends:
Figure 1: Digital transformation of securities firms
Trends
The business focus will be adjusted to achieve differentiated competition through wealth management business innovation. On the one hand, the homogeneous competition among securities companies and brokerage businesses is fierce, and the growth space is compressed and encounters bottlenecks; on the other hand, as residents’ income increases, the demand for wealth management is increasing day by day. Wealth management has become a consensus in the transformation of securities business. Leading securities firms in the industry have tried to enhance wealth management through digital means such as expanding online channels, optimizing product portfolios, and improving advisory capabilities, which involves the construction of processes, organizations, systems, and other aspects.
Strengthen business support capabilities and build a business center to support business innovation. In the past, the construction of chimney-style business systems not only brought high deployment, operation and maintenance costs, but also made business departments independent of each other, making it difficult to achieve efficient collaboration and information sharing, and could not meet the needs of rapid innovation in securities business in a fiercely competitive environment. The business middle platform can provide a unified platform for various business departments, improve resource utilization, and enhance organizational operation efficiency and innovation capabilities.
Expand the application of technological innovation and accelerate the implementation of AI+ scenarios. Work processes and models that were dominated by manual work in the past are being reshaped. AI+ can effectively improve operational efficiency and become an effective starting point for cost reduction and efficiency increase in the digitalization of securities. AI+ is being widely used in multiple scenarios such as intelligent marketing, intelligent investment research, intelligent investment consulting, intelligent customer service, and intelligent operation and maintenance.
To sum up, this report selects three key markets of wealth management, business middle office and relatively mature intelligent operation and maintenance in AI applications for application practice analysis, with a view to providing securities companies with practical experience in digital transformation and helping to efficiently promote digital transformation and upgrading.
02 Wealth Management
2.1 The growth of brokerage business is sluggish, and wealth management has become a new growth driver for securities companies
Competition among securities companies is becoming increasingly fierce. As commission rates gradually approach the cost line, the growth of traditional brokerage business of securities companies is sluggish. At the same time, the growth in personal investment scale has boosted the demand for wealth management. According to McKinsey data, domestic personal financial investment will reach nearly 250 trillion yuan in 2022, becoming the world's second largest wealth management market.
In addition, policies encourage and guide securities firms to carry out wealth management transformation. Since the launch of the fund investment consulting pilot in 2019, in June this year, the "Regulations on the Investment Consulting Business of Publicly Offered Securities Investment Funds (Draft for Comment)" was released, which represents the shift of the fund investment consulting business from pilot to regular, further promoting the transformation of securities firms' wealth management. Against this background, securities firms have accelerated the pace of wealth management transformation and regarded wealth management as a new driver of future growth.
Figure 2: Multiple factors drive the transformation of securities firms’ wealth management
2.2 Brokerage firms face new challenges in wealth management. The wealth management business process is significantly different from the traditional brokerage business process. The traditional brokerage business process is dominated by "seller" thinking and emphasizes the transaction link. The entire process of wealth management is based on “buyer” thinking, emphasizing user-centeredness. "Three-point investment, seven-point consideration" continues to operate users with a full range of advisory services, including understanding customers before investment, providing professional investment advice during investment, continuous companionship after investment, and investment education services throughout the entire life cycle.
Figure 3: Comparison between traditional brokerage business process and wealth management business process
The new wealth management process also has different capabilities requirements for securities firms than in the past, and digital capabilities are especially needed to empower each link.
Take customer profiling in the customer acquisition process as an example. Traditional securities companies mostly use questionnaire surveys to identify customers' risk tolerance. This method can easily lead to a mismatch between investment behavior and actual risk preferences. In order for wealth management to achieve the purpose of helping customers achieve wealth preservation and appreciation, it first needs to have a comprehensive and in-depth understanding of customer needs. This requires securities firms, in addition to obtaining information directly through questionnaires and communication, to integrate customers’ asset data across multiple channels and platforms to establish accurate customer portraits.
In the process of obtaining user data across platforms, privacy computing is the first choice without infringing on user privacy. Taking a leading securities firm as an example, based on multi-party joint modeling of privacy computing, multi-dimensional data was introduced in a legal and compliant manner, and an intelligent system covering more than 1,000 feature label information including customer transaction behavior, financial products, platform behavior, and demographic attributes was developed. Image tagging system. This brokerage APP can accurately assess investors' risk tolerance through user profiling and data mining, provide appropriate investment portfolio products, and push appropriate information.
Figure 4: Wealth management business scenario
Investment education also needs to use live video broadcasts, intelligent customer service and other methods to provide customers with full-process and full-lifecycle services. With the help of various investment education service methods, securities firms can fully activate and operate customers.
For example, a leading brokerage provides "Investment Consulting Online" all-weather live broadcast, providing professional investment advice and companionship in the process. Multiple large models are integrated into the APP. Through its powerful semantic understanding capabilities, it improves the accuracy of identifying the intent of customer problems, matches more accurate wealth products, and guides customers to make scientific investments. Through the above-mentioned investment education services, securities firms have improved the activity of the APP, increased the online time of customers, and ultimately obtained a higher share of customers' asset allocation.
2.3 Large models will be implemented in all scenarios of brokerage wealth management
Being customer-centric requires securities firms to meet customers’ personalized wealth management needs. The current service model of securities companies is mainly based on professional investment advisory team services, which not only requires investment consultants to have professional knowledge and experience, but also limits the coverage of securities companies to the wealth management needs of a wider range of people. Large models can not only effectively improve the work efficiency of the investment advisory team, but also cover a wider customer base with more intelligent and personalized services.
Currently, large models are used in brokerage wealth management before, during, and after investment, including intelligent investment advisory, intelligent investment research, intelligent customer service, intelligent operation, intelligent investment education, digital human, etc. Large model applications are still mainly based on single-point experiments, and intelligent customer service is the main exploration scenario. In the future, as leading securities firms become increasingly mature in training large models in vertical fields, large models will cover all wealth management scenarios and become a strong support for leading securities firms to build differentiated wealth management competitiveness.
03 Intelligent writing center
3.1 User needs and business innovation require securities firms to collaborate with multiple departments and operate efficiently
In the fierce market competition environment, in order to meet user experience needs and continue to promote innovative wealth management businesses, securities firms need to improve operational efficiency and achieve agile organizational responses to the market and needs. However, the independent business processes and fragmented information islands of multiple departments of securities firms have become obstacles to improving operational efficiency. How to use limited IT budgets to empower business systems with technical means and generate maximum benefits and value has become the focus of securities firms. Specifically, the challenges faced by securities firms include:
First, the issue of business synergy. Brokerages have multiple business departments, including IT departments, investment banking departments, wealth departments, etc. Under the traditional IT structure, each department is highly independent, and the information system is severely fragmented, making it impossible to achieve efficient collaboration and sharing. Under the current trend of increasingly fierce competition in the industry, it has created considerable obstacles for securities companies to improve their service capabilities and comprehensive business levels.
Second, there is the issue of compliance assurance. For a long time, securities have been a highly regulated industry with extremely high requirements in terms of business compliance. However, under the traditional IT architecture, due to the lack of a complete user management mechanism and content review system, securities companies lack effective technical guarantees for compliance during the business development process, which has led to many security compliance risks.
Third, there is the issue of system construction cost. Under the traditional IT system, each business department of a securities firm generally adopts a chimney-style approach to build its own IT system. Although this method can better meet the personalized IT needs of each department, it has problems such as high deployment costs, difficult operation and maintenance, and low resource utilization efficiency. It is no longer applicable in the current stage of downward revenue growth in the securities industry.
3.2 Reshape the IT system with the business center as the core and enhance business support capabilities
Based on the above background, in the face of many obstacles and problems existing in the current business operations and collaboration of securities companies, how to build an efficient, collaborative, and sustainable business system is an important issue that securities companies need to solve. Building a business middle platform is one of the effective ways to solve this problem.
First, the business center can optimize the business synergy of securities firms. The business middle platform can efficiently integrate and improve the interaction and collaboration between various business departments within securities firms through data integration, common component construction, standardized business processes, etc., thereby helping securities firms effectively get rid of the traditional business island state and achieve high efficiency, Flexible business operations improve overall business service levels.
Second, the business center can reduce the costs and risks of securities firms. By building a business middle platform, securities companies can effectively avoid duplication of technology investment, realize resource sharing and collaborative management between various departments, and unified operation and maintenance of IT facilities, thereby significantly reducing IT costs and risks.
Third, the business middle office can effectively ensure the compliance of securities firms’ business. On the one hand, through the construction of the business middle platform, securities firms can uniformly manage various business data and transaction information, and classify and process them according to different compliance requirements, thereby reducing the risk of illegal operations; on the other hand, the business middle platform can also Realize automated risk control and improve the risk control level of securities companies. By integrating the risk control model into the business center, it is possible to automatically identify, analyze and warn system risk events, thereby minimizing the probability and impact of risk events.
3.3 Steady progress from easy to difficult, and the intelligent writing middle platform empowers the business middle platform to achieve smooth implementation
To sum up, it is not difficult to see that reshaping IT and business systems with building a business middle platform as the core is becoming one of the key solutions for securities companies to continue to deepen their digital transformation. By building a business middle platform, strengthening internal business collaboration, improving competitiveness and efficiency, and constantly exploring the applications and scenarios of emerging technologies, it has gradually become a core means for securities companies to gain market competitive advantages in the future.
However, judging from the current actual progress, the business center still faces many challenges such as technical architecture compatibility and construction costs if it wants to be fully implemented within securities firms. There is still a long way to go to achieve overall large-scale application implementation. But starting from a single scene and then gradually forming full scene coverage may be a feasible path. Among them, the document writing scenario has gradually become the first step for many securities companies to build a business middle office due to its large coverage and core reasons such as not affecting business continuity.
Figure 5: Text writing scenario in the securities field
Document writing exists in the daily work of each business department of a securities firm and is a necessary foundation for the securities firm to continue to conduct business. From simple daily and weekly reports to the writing of more professional and in-depth analysis reports, employees of securities firms under the traditional system have to spend a lot of time and energy every day. However, in terms of actual results, the overall reporting output efficiency is low, and under the current trend of increasingly stringent regulatory compliance, the risk of content violations is prone to occur. Therefore, it has become an inevitable choice for securities companies to deepen their digital transformation by building an intelligent writing platform to reshape the overall document writing system, release employees' greater productivity while ensuring business compliance, and thereby empower business growth.
The intelligent writing platform refers to the use of technologies such as big data, intelligent enhanced analysis, and artificial intelligence to obtain data on demand from multi-source heterogeneous data sources, then complete the inference and processing of the original structured data according to a specific rule system, and form an analysis conclusion. or text content, thereby realizing a technical platform for automating data analysis and document writing. In the securities industry, intelligent writing platforms can help securities companies' internal investment banking, asset management, investment research, operations management, compliance management, quality control and other departments to quickly and accurately generate rich types of text content, thereby improving the efficiency of securities companies. service quality and efficiency. At the same time, relying on the multiple core advantages and capabilities of the intelligent writing platform, securities companies can comprehensively make up for the core issues and shortcomings such as synergy, compliance, cost, etc. arising from the current business and technology integration stage, as follows:
In terms of optimizing business collaboration. First, the intelligent writing center can optimize the overall business process of securities firms through automated process design and task allocation, and provide valuable support for business decisions;
Secondly, the intelligent writing center can also greatly improve the efficiency of information transmission and sharing in various business departments by integrating information from different teams and departments and sharing it with all relevant personnel;
Finally, based on the multi-person online collaboration capabilities of the intelligent writing platform, communication capabilities between different business departments can be strengthened through online editing, real-time sharing and other operations, and the collaboration of the overall business system can be further improved.
in effectively reducing costs. The intelligent writing platform is designed and built with the concept of "middle platform and modularization". Through one platform construction, it can cover all business scenarios. Based on the advantages of modular design, different functions can be developed, tested and deployed independently, thereby significantly reducing system construction costs and risks.
On the other hand, traditional text writing work often requires a lot of human resources, and the intelligent writing center can standardize and automate this process, greatly improve writing efficiency and quality, and significantly reduce labor cost investment. At the same time, the intelligent writing center can also realize data sharing and collaborative management between different teams, and also effectively reduce the cost of repeated communication between departments.
In terms of ensuring business compliance. The intelligent writing platform can use financial business rules as the basis for analysis, integrate business rules from external supervision, internal risk control/compliance/quality control, etc., and apply them to document writing and business review processes, and analyze external supervision rules Achieve real-time and dynamic response to internal risk control and compliance requirements to ensure the accuracy and compliance of the automatic processing process. Under such an operating mechanism, automation can help securities firm employees reduce their workload while also significantly reducing manual intervention, thereby effectively avoiding the risk of violations caused by human errors and omissions, and achieving the purpose of improving financial business quality control.
Not only that, the intelligent writing platform also has two core capabilities: complete data management, calling and analysis, which can effectively solve the problems of scattered data sources, low calling efficiency, and heavy workload of objective analysis under the original document writing system, and further utilize the data. Effectiveness, improve the quality and efficiency of report output.
Unified data calling ability: Under the traditional document writing system, data sources are scattered. Brokerage employees need to search for the required data sources on demand, download them, and then go through tedious processes such as cleaning before applying them to report writing. The overall process is repetitive. It is highly sensitive and takes up a lot of working time. The intelligent writing center can automatically complete the docking with multiple internal and external data sources of securities companies through the processing of the rules engine, automatically call the data required by employees with one click, and automatically clean, deduplicate and cross the data during the call process. Checksum accuracy verification. After these automatically obtained multi-source data are automatically inspected by the system, analyzed and verified by business personnel, reviewed by the audit department, and even submitted and disclosed to the outside world, they can form their own trusted business database within the business department or securities company, breaking the existing The data island system improves data application efficiency. There are two core steps in building a data middle platform: "use" and "storage". Through intelligent writing, the middle platform not only realizes the flexible call of multi-source data, but also completes the data accuracy and validity confirmation through the "use" link, so that it can "Cun" is the company's data asset, truly realizing the interconnection between data and business, which also lays a solid foundation for building a data center.
Multi-dimensional data analysis capabilities: Report writing in the securities field requires high professionalism and often involves a large amount of objective data analysis work. Under the traditional system, these tasks are also completed manually, resulting in subjective analysis time being occupied and the overall report output Quality suffers. Most intelligent writing platforms have built-in rich financial business analysis functions and analysis modules, which can automate most of the objective analysis work, leaving more time for subjective analysis, greatly improving employee productivity and report output quality.
To sum up, with the development of financial technology and policy support, the digital transformation of securities firms is gradually deepening, and the intelligent writing platform, as an important component of digital transformation, can help securities firms realize automatic generation, processing and management of text information, thereby better It can fully adapt to the business needs of the digital era, achieve comprehensive optimization of business processes, and significantly improve service quality and overall industry competitiveness.
Typical Case 1: Relying on intelligent core technology, China International Finance Securities uses Yuxin Intelligent Writing Center to improve multiple business capabilities
Founded in 1990, Sinolink Securities is a listed securities company with excellent asset quality, a capable professional team, and outstanding innovation capabilities. In recent years, under the strong advocacy of the digital transformation of national financial institutions, Sinolink Securities has built a first-class securities technology organization with the core goal of "integrating businesses and empowering platforms", continues to deepen the integration of technology and business, and is committed to using technology to create the largest business Value, has now become one of the more successful securities companies in digital transformation; Shenzhen Yuxin Technology Co., Ltd. is a "Municipal Innovative Small and Medium-sized Enterprise" and a "National High-tech Enterprise". Relying on its understanding of financial business, Yuxin Technology is committed to providing professional-level financial business digital solutions for financial institutions with three major implementation scenarios: intelligent financial business analysis, automated financial document writing, and digital business quality control and review. , has now been implemented in more than 200 financial scenarios in investment banking, asset management, brokerage, investment research/research institutes, IT operations, credit, project quality control, risk control, and compliance in securities firms, banks, insurance, funds, and other institutions.
1. Digital transformation continues to deepen, and securities companies have discovered four productivity improvement opportunities under the "three typical industry exhibition scenarios"
In order to seize the transformation and upgrading opportunities brought by the digital wave of the securities industry, the management of China International Finance Securities based on the actual situation of the company, and jointly reviewed the current situation and business of China International Finance Securities with Yuxin Technology and found that various departments of China International Finance Securities are in daily work. All involve a large amount of "document writing", "uploading and distributing", and "summary analysis" work. These three major industry exhibition scenarios all use electronic documents as information transmission carriers, including daily reports, weekly reports, various contract documents, marketing materials, and investment banking business materials. , research reports, compliance analysis reports, risk reports, management reports, etc. As the business scale and market share of Sinolink Securities continue to increase, the number of the above types of documents that need to be written in daily work has also increased exponentially. In some departments, the above three major exhibition scenarios account for more than 50% of the daily workload. . In the past, when faced with the need for massive document generation and summary management in the securities industry, business colleagues would use electronic document management tools, but most of the work was still done manually. After careful analysis and sorting out, Sinolink Securities and Yuxin Technology came up with the following four digital opportunities. Ai Analysis believes that these four points are common among domestic securities companies of a certain size. It is organized as follows:
1.1 Digitization of project management and control: releasing managers’ time and energy
Take the daily management of the IT department of Guojin Securities as an example. In order to facilitate the upper management to better control the progress of the project, during the daily operations of Sinolink Securities, IT department employees must regularly write daily and weekly reports and record project progress in real time. Under the traditional document management system, employees submit daily and weekly reports to managers through traditional methods such as IM and email. After receiving relevant documents, managers need to spend time manually organizing, summarizing, reading, and energy to extract and discover key information. , and finally gave management actions; on the other hand, in recent years, the management of China International Finance Securities has followed the trend and continued to increase investment in digitalization. This has also been accompanied by the continued expansion of the scale of the financial technology team, and the workload of reading and replying to traditional daily and weekly reports has increased rapidly. , which greatly increases the burden of analysis and decision-making on managers, and is not conducive to timely and reasonable control of project progress and risk management. Sinolink Securities The rapid development of financial technology has brought about new management problems that are common among various securities companies and can be properly solved by relying on digital means.
1.2 Automation of text writing: common and high-frequency scenarios, releasing massive productivity and improving compliance
It is common in securities companies that employees in various departments need to write a lot of text reports every day. In essence, they complete the specific execution work of various financial businesses through document writing, such as data acquisition, business analysis, rule reasoning, uploading and issuing, external Submissions, etc. to advance the business process. In the current traditional document writing method where people use computers, tablets and other tools to complete electronic document writing, typesetting, proofreading and delivery, employees in almost every department need to spend a considerable proportion of time and energy completing these electronic document writing tasks. ; Not to mention that in the financial field with strong supervision, various compliance risks often occur due to human errors and omissions in document writing and business analysis, which may affect business progress and company reputation in the slightest, or subject to regulatory penalties or even Business is suspended. On the one hand, it is time-consuming, cumbersome, and error-prone; on the other hand, it is high-frequency and common. This contains huge potential data opportunities.
1.3 Build an effective cycle of data "storage" and "use": break data silos and let data flow
Different from ordinary writing, various texts written by various business departments of securities companies basically require a large amount of underlying data to support them. The rigor of financial business determines the high requirements for the accuracy and reliability of data. Usually, the data is input into various texts only after the authenticity of the data has been confirmed by the experience of business personnel. On the other hand, the phenomenon of "data islands" is prevalent in the securities industry. There is no one way to organically combine the internal systems and external data sources of various securities firms to form a reasonable business empowerment. In the actual business development process, financial practitioners need to obtain different data and information from different websites, data terminals, information channels, or (financial) project parties based on their experience, and manually correct and confirm the authenticity of the data itself and data sources. , and then after the data is analyzed and processed based on the regulatory requirements and business rules of the business, the analyzed results can be manually input into the report. How to break data silos and free employees from the current tedious work of "finding data"; how to connect data and business through "using data" so that the basis of digitalization - huge amounts of data cleaning and integration work can be integrated into business Continuously and efficiently; how to automatically and accurately "save" data after completing business goals and establish the data assets of securities companies are all digital issues worth exploring now. After all, the value of big data does not lie in the huge amount of data, but in It lies in establishing a data-based business cycle.
1.4 Digital intelligence in financial business analysis: “Leave objective analysis to machines” and integrate digital technology into business
Document writing in financial scenarios is different from the writing of novels, essays, poems, and argumentative essays. The writing process has the characteristics of multiple sources and complex analysis. It requires a high degree of specialization and requires business personnel to invest a lot of time and energy. Under the traditional document writing system, text writing often has "rules to follow", and the existence of this "rule" means that financial text writing needs to follow its own unique set of financial rules. This rule system has the following two characteristics: first, multiple sources. Rules may come from laws and regulations, regulatory department requirements, or the company's internal risk control and compliance management requirements. In addition, there are a large number of financial business rules that also come from financial knowledge in professional fields. Second, although the rule system comes from many sources, is highly professional, and is dynamic and changeable, the entire rule system is still relatively fixed. It can automatically call, reason, analyze and output results after modeling the rule system through digital means. From a digital perspective, these objective analysis tasks that need to be repeated and take up a lot of working time can be completed by a digital system; by replacing this part of the work, financial practitioners can have more time. And energy is used for more productive work such as in-depth analysis, subjective judgment, customer development, etc., to improve overall business efficiency.
2. Comprehensive upgrade of technical capabilities, Sinolink Securities uses Yuxin Intelligent Writing Center to reshape the document writing system
Based on the above analysis, it can be seen that among the many business departments and business processes of securities companies, only in the scenario related to "document writing", through reasonable and innovative digital transformation, it can achieve the goals of data asset precipitation, business process reengineering, productivity release, and cooperation. Gain benefits in many aspects such as regulatory quality improvement. The ubiquity and high frequency of document writing-related scenarios within various financial institutions mean that digital construction in this scenario will have better input and output.
Based on the exploration and sorting out of the above-mentioned digital opportunities, in order to continue to deepen the digital transformation process, Sinolink Securities decided to carry out digital construction to seize "document writing", a typical scenario for securities companies. Utilize a variety of advanced technologies such as artificial intelligence and big data to build an intelligent document writing center to serve multiple business departments to the greatest extent, reduce manual intervention links, improve document output efficiency, and greatly optimize employee experience and improve quality. control level, thereby enabling sustained and healthy growth of the business.
In the product selection stage, affected by external factors such as the strengthening of financial industry supervision in recent years and the credit innovation policy, Sinolink Securities conducted a focused inspection of the supplier's qualifications, business compliance, and securities industry service experience. In the end, after multi-party research and product POC testing, Shenzhen Yuxin Technology Co., Ltd. relied on its deep understanding of the securities business, comprehensive leading product performance, relatively complete solutions, and rich actual implementation cases of leading securities companies to become product supplier for this project. In response to the various needs of China International Finance Securities for intelligent document writing platforms at this stage, Yuxin Technology, after accurately understanding the business goals of China International Finance Securities, has output a set of intelligent text generation, automatic reasoning of business rules, and An intelligent writing middle-end solution with core capabilities such as a rich-format editing interface and a visual configuration management backend.
Figure 6: Schematic diagram of Yuxin Technology’s intelligent writing and middle-end capabilities
During the project deployment phase, in order to meet the needs of the Guojin Securities system to be quickly launched, with the help of Yuxin Technology's product-friendly technical architecture, Guojin Securities and Yuxin Technology completed the platform deployment in just one week, and in the Science and Technology R&D Department , the Information Technology Department was the first to put it into use, using the multiple core capabilities of the intelligent writing platform to upgrade the original system upgrade requirements one by one, and relying on the built platform-level capabilities to extend the intelligent text writing capabilities to multiple departments throughout the company Promote it to maximize the digital benefits of the middle platform. The specific advancement path is as follows:
Figure 7: China International Finance Securities launches smart writing platform in stages
2.1 Complete text summary and management mechanism to significantly improve the company’s business management and control level
In response to the problems existing in the document summary management system of the IT department in the industry, on the one hand, Sinolink Securities uses the Yuxin Intelligent Writing Center to provide it with unified and automatic document summarization capabilities, which can automatically summarize the documents written by the IT department employees of Sinolink Securities every day. The daily and weekly reports are summarized in a unified manner and sent to the relevant managers’ emails, which greatly simplifies the summary process;
On the other hand, Sinolink Securities uses the intelligent writing platform to conduct summary statistics and basic analysis of important indicator data in daily and weekly reports according to the needs of managers, so that managers can quickly discover project risks without reviewing the contents of employee reports one by one. It comprehensively optimizes the traditional project management model and improves managers' project control capabilities.
2.2 Automation + intelligent writing capabilities help employees comprehensively improve their text writing efficiency and experience
Yuxin Technology has been deeply involved in the securities industry for many years and has accumulated a deep understanding of industry know-how and business scenarios. Its intelligent writing platform has built-in 200+ document templates that can be applied to different business needs. At the same time, the intelligent writing platform also has mature low-code capabilities, allowing business personnel to customize and develop new text templates to further meet the personalized text writing needs of various business departments.
Based on a large number of text templates, combined with the powerful personalized configuration capabilities of the middle platform and intelligent enhanced analysis, data intelligence and other technologies, through the automatic acquisition and filling of multiple data sources, Sinolink Securities uses the intelligent writing middle platform to enable employees in various departments to not only quickly Generating the document reports you need eliminates the need to perform a large amount of repetitive work every day. It also significantly reduces the project quality risks that frequently occur due to manual errors and omissions in the financial development process, and improves the level of financial quality control.
2.3 Flexible and diverse internal and external data source docking methods to meet the unified collection and call requirements of all data
In order to effectively solve the common problems in the traditional text writing system of business departments, such as scattered data sources and difficulty in collecting and calling underlying data, Sinolink Securities made good use of the unified data collection and calling capabilities of Yuxin Intelligent Writing Center. Based on the multiple flexible docking methods of this intelligent writing platform, it can realize seamless docking with various internal databases and RPA tools of China International Finance Securities, as well as with various external data sources, so that the fragmentation between various data sources can be solved. Great improvement, the value of data can also be better utilized. At the same time, employees no longer need to go through tedious data search and download processes when writing daily texts. They can automatically obtain and call multiple data sources in real time with one click, which greatly optimizes the efficiency and experience of employees' daily text writing.
2.4 Industry-leading multi-dimensional data analysis capabilities, Sinolink Securities fully unleashes employee capabilities and value through the intelligent writing platform
In terms of objective business analysis needs, Sinolink Securities has obtained mature multi-dimensional data analysis capabilities by using Yuxin Intelligent Writing Center. Based on the industry-leading financial business rules engine of this intelligent writing platform, as well as more than 8,000 financial rule inference functions and more than 1,000 financial business analysis modules, it can automatically, compliantly and quickly output objective analysis results according to user requirements. Assisted the writing of various in-depth analysis reports for different business departments of Sinolink Securities, truly achieving the remarkable effect of “leaving objective analysis to machines and leaving subjective judgment to humans”.
- Sinolink Securities has fully implemented the intelligent writing platform to realize a variety of values and benefits.
Text writing efficiency has been comprehensively improved. Through the overall deployment and application of Yuxin's intelligent writing platform, Sinolink Securities has realized the reshaping and upgrading of the text writing system and completed a huge transformation from traditional manual writing to intelligent and automated writing. On the premise of ensuring business compliance, text writing work that traditionally took several days to complete can now be completed in just "a few minutes," greatly improving text writing efficiency.
The value of employees has been further developed. Sinolink Securities relies on Yuxin Technology’s sufficient business understanding of the securities industry and the various automated and intelligent functions of the intelligent writing platform to fully optimize the employee document writing experience. On the premise of ensuring business compliance, employees do not need to By carrying out a large amount of repetitive work such as data collection and objective analysis, you can devote your limited energy to more important business and further develop your own value and capabilities.
So far, Sinolink Securities has completed the launch and use of the intelligent credit writing platform, and has gradually rolled it out to various business scenarios, helping Sinolink Securities integrate the external regulatory framework, internal risk control requirements, financial expertise, and daily business operations. These complex professional-level rules are integrated and applied through reasonable technical means, thereby realizing software replacement for financial analysis work and financial document writing work, and releasing considerable productivity. In the future, the cooperation between the two parties will be further deepened to jointly explore more application scenarios of artificial intelligence technology and provide assistance for the comprehensive acceleration of the digital transformation of Sinolink Securities.
3.4 Under the wave of generative AI, intelligent writing platform is an important application scenario of large models
Since 2022, AIGC and large model technology have gradually come into people's vision, and their powerful content production capabilities have brought great shock to people. Similarly, for the securities industry, the arrival of large model technology has also brought new opportunities for securities firms to achieve an overall improvement in business productivity. However, judging from the current actual implementation progress, on the one hand, the regulatory compliance requirements of the securities industry are strict, and the compliance of large model algorithms cannot be guaranteed. On the other hand, the business in the financial industry is highly professional, and current large models lack sufficient business understanding and industry data, which are far from the actual business needs of the securities industry. Therefore, based on the above two core difficulties, if large models want to be quickly implemented in the securities industry, combining them with intelligent writing platforms may be a feasible path. The specific reasons are as follows:
Figure 8: Intelligent writing of architecture diagram combining middle platform and large model
First, if a large model is to become a large industry model applicable to the securities industry, it requires a lot of fine-tuning of prompts and instructions. The massive financial business rule functions built into the bottom layer of the intelligent writing platform can just solve the problem that the current large model lacks industry data and cannot be close to the actual business needs.
Second, for compliance issues, the product features of the intelligent writing platform that have integrated external supervision, internal risk control/compliance/quality control and other business rules can fully ensure the compliance of the content generated by large models. Not only that, for the intelligent writing platform itself, combining with large models is also an important step to achieve further intelligent upgrades.
Currently, the intelligent writing platform still requires manual configuration of templates and other tasks. After combined with the large model, securities firm employees can generate documents only through a conversational method, truly realizing the intelligent upgrade and reshaping of the business model and liberating maximum productivity. In the future, with the comprehensive deepening of the digital transformation of securities, with the support of advanced technologies such as large models and RPA, the intelligent writing platform can further improve the accuracy and efficiency of intelligent writing, reduce costs, and bring more value to the securities business. By combining with other technologies, the intelligent writing platform will also be further extended to more business scenarios, truly helping to accelerate the overall digital transformation of securities.
04 Intelligent Operation and Maintenance
4.1 The transition of securities firm IT operation and maintenance from platform-based operation and maintenance to intelligent operation and maintenance
The IT operation and maintenance of securities firms continues to evolve with the penetration of digitalization, from the initial manual operation and maintenance stage to the current platform-based operation and maintenance stage. Most securities firms have built their own operation and maintenance platforms, integrating monitoring platforms, alarm platforms, automation platforms and other tools and systems to improve the stability of production systems through platform-based operation and maintenance. However, in the platform-based operation and maintenance stage, operation and maintenance still faces problems such as poor data quality, low data timeliness, and a large amount of repetitive work, and operation and maintenance efficiency needs to be improved. Fierce competition in the market has driven securities firms to accelerate business innovation, such as exploring new wealth management businesses and expanding Internet channels. Business innovation has made brokerage business systems more complex, which will inevitably lead to an increase in the complexity of the operation and maintenance system. With the addition of new technology applications such as cloud native, the issues that operation and maintenance personnel need to pay attention to increase exponentially, bringing new challenges to operation and maintenance work. Operation and maintenance personnel need to have the ability to analyze and predict massive events on the basis of supervising and controlling business systems. For example, analysis of massive alarm information, rapid location of fault causes, efficient log analysis, etc. In order to solve the above problems, leading securities firms have enhanced their analysis capabilities through intelligent operation and maintenance. Intelligent operation and maintenance uses AI algorithms to analyze massive data to achieve early warning of faults, denoise, aggregate and classify alarm events, and complete rapid root cause location.
4.2 Conditions for securities firms to realize intelligent operation and maintenance
In order to realize intelligent operation and maintenance, securities firms need to have two conditions: data management and intelligent computing engine.
Figure 9: Prerequisites for securities firms to realize intelligent operation and maintenance
First, data management needs to be implemented. Gather, manage and analyze structured and unstructured data from a variety of monitoring and operation and maintenance tools. In particular, it is necessary to achieve real-time and quasi-real-time calculations of massive data to meet the timeliness of operation and maintenance scenarios such as fault prediction. Require. Secondly, it needs to have an intelligent computing engine with high-quality algorithms for specific scenarios, such as anomaly detection, alarm convergence, fault prediction, resource optimization, etc. After achieving the above conditions, securities companies need to combine business pain points and implement them in specific scenarios such as failure, quality, safety, efficiency and performance. To sum up, intelligent operation and maintenance is a new stage in the evolution of IT operation and maintenance of securities companies. Intelligent operation and maintenance will also be an important part of the digital transformation of securities companies and promote the rapid development of securities companies' business.
Typical case 2: Intelligent operation and maintenance analysis platform accelerates the evolution of Essence Securities’ intelligent operation and maintenance analysis capabilities
Essence Securities Co., Ltd. is a fully licensed comprehensive securities firm with many business rankings among the top in the country. It is headquartered in Shenzhen and has 50 branches in Beijing, Shanghai, Guangzhou, Shantou, Foshan and other places. In 29 There are 320 securities business offices in each provincial administrative region. Essence Securities takes "financial services to create a better life" as its mission and strives to become a highly respected first-class capital market service provider.
The securities industry has strict requirements on system continuity and stability. In the securities industry, securities investment information changes rapidly, and customer interests are of paramount importance. Ensuring the real-time, availability, and stability of IT operation and maintenance systems to avoid accidents such as data disorder, data loss, and transaction interruption are extremely important to the daily operations of securities companies. It's important. At the same time, supervision has strengthened requirements for the continuity of securities services, and the "Guidelines for the Continuity Management of Information Technology Services in the Securities and Futures Industry" will be released in 2022 to provide guidance on the basic procedures and measures for the continuity management of information technology services for securities companies. In the era of digital transformation, technological innovation has become the core driving force for the development of the financial industry.
Essence Securities adheres to the corporate vision of becoming a first-class capital market service provider with the most market value and core competitiveness in China and is widely respected. It has always followed the pace of technological innovation and development, adhered to the promotion of digital transformation strategy, and continued to optimize to achieve technological empowerment. Digital infrastructure and basic operation support platform. As early as 2020, Essence Securities has built an operation and maintenance monitoring platform, and continues to optimize and improve functions, moving towards an integrated intelligent operation and maintenance platform. In accordance with the plan of the intelligent operation and maintenance integrated platform, Essence Securities has completed the first phase of "platform construction" and the second phase of "capacity building" to create an integrated operation and maintenance tool platform system architecture with low-code operation and maintenance development capabilities. Establish a "supervision and control" process for IT operation and maintenance.
Manual-based operation and maintenance model faces efficiency challenges
With the rapid development of Essence Securities' business, the growth of system complexity has brought about a rapid expansion of the volume of operation and maintenance data. IT operation and maintenance personnel are faced with the analysis and processing of massive data. The existing operation and maintenance model is mainly based on manual analysis and is passive to risks. Response, resulting in low operation and maintenance efficiency and fluctuation in operation and maintenance quality, making it difficult to support Anxin's business innovation and iteration and ensure user experience. How to improve IT operation and maintenance analysis capabilities and efficiency is an important challenge currently faced by Essence Securities. The specific manifestations are as follows:
Operation and maintenance data has not been uniformly collected, which affects the efficiency of troubleshooting. In securities companies, the storage paths, formats, specifications, etc. of operation and maintenance data vary and are scattered in various systems. Essence Securities has centralized storage and management of important data from most monitoring tools, such as indicator data, log data, alarm data, etc., and has implemented unified storage and established a set of operation and maintenance data standards. However, non-monitoring data such as CMDB data, work order data, etc. have not yet been integrated, and their access needs to span multiple systems and tools, making the operation and maintenance data search path complicated and increasing troubleshooting time.
Traditional IT operation and maintenance relies heavily on the experience of operation and maintenance personnel, which increases operation and maintenance costs. For example, in alarm management, alarms in various monitoring systems of Essence Securities need to be implemented by setting static threshold rules. The setting of rules has a high threshold for operation and maintenance personnel, requiring operation and maintenance personnel to have rich experience accumulation. Moreover, because threshold rules are affected by the operating system, once the operating system changes due to factors such as business model, load model, user behavior, etc., the threshold rules need to be reset, which makes the maintenance and management of threshold rules also consume a lot of manpower. In addition to alarms, log data comes from operating systems, applications, network devices, etc. The amount of data is huge and complex. In situations such as correlation analysis of log events, troubleshooting, and abnormal pattern recognition, operation and maintenance personnel are required to have strong skills. Technical knowledge and rich experience for analysis and processing.
The quality and quantity of intelligent operation and maintenance algorithms cannot meet the needs of intelligent operation and maintenance scenarios. For example, the accuracy of identifying alarm noise in Essence Securities’ smart alarm is low, and the operation and maintenance team still needs to spend a lot of energy to deal with noise alarms. For business personalized intelligent operation and maintenance scenarios, such as intelligent log comparison and alarm correlation analysis in application change scenarios, the current algorithm capabilities cannot support the implementation of the scenario.
Lack of observability and management analysis capabilities for the overall IT operation and maintenance health of the business system. When understanding the overall situation of the business system, such as performance, collaboration, and resources, since the operation and maintenance platforms within Anxing Securities are independent of each other, operation and maintenance personnel need to switch to multiple platforms to collect information, which is inefficient. Especially when troubleshooting business system failures, the operation and maintenance team needs to troubleshoot multiple systems at the same time, affecting system continuity.
In order to solve the above pain points, Essence Securities IT operation and maintenance analysis needs to change the operation and maintenance model, from a manual-based and passive IT operation and maintenance model to a more efficient and intelligent active operation and maintenance model. , and clarified that the third stage of the intelligent operation and maintenance integrated platform is "scenario construction", and the platform construction evolves to "scenario and digital driven". Essence Securities will also make full use of AI and other technologies to empower business scenarios, explore and practice intelligent operation and maintenance scenarios, and incorporate them into one of the strategic measures to achieve comprehensive digital transformation during the "14th Five-Year Plan" period. After multiple investigations and comprehensive consideration of various factors such as the manufacturer's technical capabilities, service capabilities, successful cases, and project delivery capabilities, Qingchuang Technology finally relied on its advanced technology, mature platform products, continuously evolving platform expansion capabilities, and powerful comprehensive Strength and other core advantages, after going through POC, bidding and other processes, it has become Essence Securities’ partner for the implementation of intelligent operation and maintenance scenarios.
Shanghai Qingchuang Information Technology Co., Ltd. ("Qingchuang Technology" for short) is a provider of intelligent operation and maintenance AIOps solutions for all areas. It focuses on empowering operation and maintenance management with AI, activating operation and maintenance data, and optimizing operation and maintenance efficiency. At present, customers have covered multiple industries such as banking, insurance, securities, manufacturing, energy and transportation. Its self-developed Sherlock AIOps smart operation platform integrates multi-dimensional data such as alarm events, performance indicators, logs and capacity, and combines it with AI algorithm technology to provide operation and maintenance solutions including precise alarms, anomaly detection, root cause location, etc., through the value of data. Refining analysis and optimizing business operation decisions. And Sherlock AIOps fully supports the Xinchuang adaptation of domestic software and hardware products such as databases, middleware, cloud platforms, chips, and operating systems.
Starting from the business perspective, Essence Securities continues to optimize its intelligent operation and maintenance analysis capabilities
The cooperation between Essence Securities and Qingchuang Technology has gone through two phases. In the first phase, the construction of the intelligent operation and maintenance analysis platform was completed, and functions such as single indicator anomaly detection, log anomaly detection, and alarm compression were completed; in the second phase, the intelligent operation and maintenance analysis platform was developed horizontally. Expand to cover more business systems, and develop personalized intelligent operation and maintenance scenarios based on the specific demands of business systems, such as intelligent comparison of logs, recommendation of fault root causes, etc. The specific solutions are as follows:
Figure 10: Essence Securities Intelligent Operation and Maintenance Analysis Platform Architecture Diagram
1. Build an intelligent operation and maintenance middle platform, uniformly store and efficiently manage operation and maintenance data, and provide algorithm and data support for intelligent operation and maintenance scenarios.
With the support of Qingchuang Technology, Essence Securities first integrated multi-source operation and maintenance data and adopted the Clickhouse low-cost storage solution to achieve rapid access and unified management of multiple operation and maintenance data such as Zabbix, Promethues, BPC, and log platforms. Afterwards, operation and maintenance personnel can convert and clean multi-source data graphically on the data governance platform to improve data quality. The algorithm platform supports intelligent operation and maintenance. In the algorithm platform, a variety of algorithms that Qingchuang Technology can use out of the box are embedded, such as indicator anomaly detection, log anomaly detection algorithm, trend prediction algorithm, clustering algorithm, etc., and supports operator binary expansion, so operation and maintenance personnel can Independently model and adjust algorithm parameters to empower intelligent operation and maintenance scenarios.
2. Build intelligent operation and maintenance capabilities through intelligent log analysis and intelligent alarm analysis to assist fault location and root cause analysis scenarios
In terms of intelligent log analysis, intelligent log comparison effectively supports application change scenarios. Intelligent log comparison can perform fully automatic verification and analysis of log data before and after system changes, discovering deviations that are difficult to detect manually, such as "novel" templates or variables, complementing the blind spots of traditional monitoring and testing methods, eliminating potential hazards, and ensuring system changes After going online, the business can run normally. Log feature detection provides log anomaly detection and provides intelligent operation and maintenance analysis for operation and maintenance personnel. Combined with intelligent algorithms to conduct real-time analysis of logs, automatically identify the similarity of logs, aggregate data with high similarity, extract common log patterns, quickly grasp the full picture of logs, cluster massive logs into a number that is readable by the naked eye, and Intelligently identify the rules of log occurrence, automatically identify abnormal events such as system crashes, performance degradation, security risks, etc., provide intelligent alarms, and provide support for rapid root cause location.
Figure 11: Schematic diagram of intelligent log analysis
In terms of intelligent alarm analysis, the intelligent operation and maintenance analysis platform provides intelligent alarm compression function. It automatically clusters alarm content through text clustering algorithm and groups and compresses it according to key fields to reduce alarm notifications, improve alarm quality and readability, and Reduce the complexity of operation and maintenance configuration rules. Alarm correlation scenario analysis is based on the correlation analysis algorithm FP-Growth for data modeling, which can perform noise reduction and correlation analysis on massive alarm events, and assist in the root cause location alarm analysis and disposal process. Relevant scenario analysis also supports the accumulation of fault handling knowledge, reducing duplication of operation and maintenance work and improving operation and maintenance efficiency.
Figure 12: Schematic diagram of intelligent alarm engine
3. Establish unified monitoring capabilities through unified search and system portraits
Unified search allows operation and maintenance personnel to perform one-click queries on indicators, alarm logs, architecture diagrams, change plans and other data of a certain host within a certain time period on one page, assisting rapid emergency troubleshooting.
Figure 13: Schematic diagram of unified search function
The business system portrait enables observable business system operation and maintenance. The system portrait connects logs, business, CMDB and other systems, and provides performance views of the operating system, middleware, business, link tracking, etc. At the same time, monitoring performance indicators and alarms are presented in real time to help operation and maintenance personnel evaluate the operation and maintenance health of the business system. Overall assessment, once a failure occurs, quickly assess the associated impact and assist in locating the root cause of the system to improve operation and maintenance efficiency.
Figure 14: Business system portrait
4. Recommendation of root cause of failure
In addition, Essence Securities and Qingchuang Technology jointly researched fault root cause recommendations, and provided preliminary analysis and suggestions for system faults by integrating multi-source operation and maintenance data and integrating multi-modal analysis algorithms to help operation and maintenance personnel quickly identify problems.
Intelligent operation and maintenance analysis platform brings significant improvement in operation and maintenance efficiency
With the assistance of Qingchuang Technology, Essence Securities quickly launched an intelligent operation and maintenance analysis system. The intelligent operation and maintenance capabilities based on intelligent log analysis and intelligent alarm analysis effectively improve operation and maintenance efficiency and reduce operation and maintenance costs.
Intelligent alarm correlation analysis promotes proactive operation and maintenance. In the analysis of related scenarios in intelligent alarms, the platform is based on about 330,000 alarm data from the core business system in the past 6 months, combined with AI algorithms, and after 5 rounds of learning processes and 2 results review processes, 22 meaningful transactions were generated The correlation analysis results greatly improve the alarm value and alarm processing efficiency, and promote the development of traditional passive response to active operation and maintenance.
System portrait and unified search effectively improve operation and maintenance efficiency. The system portrait provides a global analysis perspective of the business system, integrates the key information of the system, and makes the health of the system clear at a glance, simplifying the process and reducing communication costs. One-stop unified retrieval helps operation and maintenance personnel to perform troubleshooting operations on one interface, and analyzes the changes before and after the fault through the timeline. Compared with traditional manual operations that require a large number of manual switching back and forth between platforms, it improves operation and maintenance work efficiency by 5 times. .
Intelligent log comparison shortens testing time and effectively avoids online risks after application development changes. Fully automated log verification and analysis is implemented in log change anomaly detection. The test time is reduced from the original 5-6 hours to 1 hour, which increases the test efficiency by at least 5 times. Essence Securities can discover 10+ potential risks through the platform every year. Effectively avoid risks.
Centralized storage of operation and maintenance data reduces operation and maintenance costs. After one year of dumping operation and maintenance data using a low-cost solution, 11 servers were reduced, and resource usage was reduced by 38% year-on-year.
Scenario-driven, data-driven and continuous operations are the keys to building intelligent operation and maintenance analysis capabilities
Looking back at the construction process of Essence Securities’ intelligent operation and maintenance analysis platform, its scenario-driven, data-driven concepts and continuous operation mechanism are the keys to building intelligent operation and maintenance analysis capabilities and continuously obtaining intelligent operation and maintenance benefits. The construction of Essence Securities' intelligent operation and maintenance analysis platform is driven by business scenarios, making the platform construction goals feasible and the platform construction effects measurable. IT supports business development, and business needs can also reflect on IT infrastructure. In the process of building an intelligent operation and maintenance analysis platform, Essence Securities has always adhered to "data-driven" and fully exploited the value of data to support intelligent application scenarios through unified storage and management of multi-source data and combined with AI algorithms. After the construction of the intelligent operation and maintenance analysis platform is completed, Essence Securities has established a continuous operation mechanism to ensure that the platform can quickly adjust and continuously optimize as the business system changes, and achieve lasting performance. For example, the business department will confirm and provide feedback on the intelligent log comparison, intelligent alarm and other information on the platform, and assist the platform operation and maintenance personnel to continuously optimize the algorithm.
4.3 Large models will be deeply integrated with intelligent operation and maintenance to expand operation and maintenance scenarios and improve operation and maintenance efficiency.
Currently, the scenarios for intelligent operation and maintenance of securities firms focus on anomaly detection, fault prediction, alarm convergence, root cause analysis and other scenarios. With the rise of large models, the integration of intelligent operation and maintenance and large models has expanded penetration scenarios to include intelligent question and answer, fault self-healing, fault prevention, and knowledge base construction.
On the one hand, large models can expand intelligent operation and maintenance scenarios and realize more scenario automation. During the automated deployment process, the large model can automatically generate scripts, configurations, tests and other related codes based on the natural language instructions of the operation and maintenance personnel. In the intelligent question and answer scenario, operation and maintenance personnel are provided with expert-level guidance and support through natural language interaction.
On the other hand, large models will improve the effectiveness of existing algorithms and further improve operation and maintenance efficiency. For example, in log analysis, large models can be analyzed through log data to achieve anomaly detection and root cause analysis.
Figure 15: Large model application direction
In the future, large models will become the core technology of intelligent operation and maintenance, promoting intelligent operation and maintenance to cover all operation and maintenance scenarios such as faults, quality, safety, efficiency and performance.
05 Conclusion
Digitalization has become the long-term development strategy of securities companies. Whether it is traditional brokerage business or new wealth management business, they all require digital empowerment. In digital construction, data capability building and large model application are important investment directions for future securities firms. Data capabilities are the foundation of business operations. At present, securities companies are accelerating the sorting out of data assets, improving data quality, and realizing unified storage and analysis of multiple heterogeneous data through data lakes, integrated lakes and warehouses and other data architectures, opening up data islands, and providing high quality for business innovation and large model applications. data. The large model will accelerate the implementation of innovative business scenarios for securities firms and improve operational efficiency. In particular, the natural language interaction capabilities of large language models will greatly improve manual efficiency and expand the space and time for securities companies to serve customers.