Interview with CTO Joshua Xiang of Yixin: Technology Should Serve Scenes, AI Is Naturally Suitable for Financial Industry


(Yixin CTO and Executive Director of Science and Technology Commission Joshua Xiang)

Introduction: With the outbreak of artificial intelligence technology, artificial intelligence has become a national strategy in 2017 and has begun to land in more and more industries. However, the traditional financial industry, impacted by the Internet technology, has also embarked on the road of transformation to digital and intelligent. InfoQ reporter interviewed CTO Joshua Xiang of Yixin to talk about the application of AI in the digital transformation of the financial industry.

“AI is more suitable for serving scenes than promoting new industries.”

Conform to the Development Wave and Realize Digital Transformation

Joshua Xiang believes that from the impact of technology on the industry, the development of the Internet has gone through three important stages.

  • Informatization: The most important sign is the appearance of PC. Many enterprises have started to use ERP, CRM and OA systems in the background, moving away from paper and manual operations to automation, in order to improve efficiency and reduce costs. Data has also begun to settle down.
  • Internalization: At this stage, enterprises can provide services to the outside world through the connection with the Internet. A typical example is e-commerce: services and merchan dise transactions are transferred from offline to online. At the same time, huge amounts of data such as commodities and user behaviors have been generated, and technologies such as search engines and recommendation algorithms have been created, enabling enterprises to use these data to help businesses and services to adjust and upgrade.
  • Intelligentization: Represented by AI and combined with various terminal devices, personalized customized services can be provided under specific scenes. With the development of AI and data science, big data is deeply involved in all aspects of enterprise operation and decision-making, and plays a key role in the enterprise development process. As a result, the industry has entered the digital wave stage.

At each stage, some enterprises can seize the technological trend at that time and achieve themselves.

  • At the beginning of the Internet, people mainly obtained information on the Internet. Sina, Netease and Sohu three major portals complied with the development of browser technology and achieved great success through the presentation of information.
  • After the emergence of search engine technology, through keyword search can get accurate matching, new business models (such as bidding ranking advertising business) came into being, spawning a number of companies such as Google and Baidu. With the development of Internet companies, online shopping has become possible. As a result, e-commerce companies such as Amazon, Ali and Jingdong have emerged. The rise of social software has led to social giants such as Tencent and Facebook.
  • In the era of intelligence, the popularity of mobile Internet technology has spawned a number of smart phone companies.

Although there have been a number of unicorn companies in the artificial intelligence technology industry, in Joshua Xiang’s view, AI technology is not the same as the above mentioned technology, it is more suitable for use in a certain scenario rather than creating a new industry, so it is difficult to create another batch of giant companies. With the current technology trend and market environment, enterprises should conform to the development trend and realize digital transformation by combining AI, big data and other technologies.

Challenges Faced by Digital Transformation of Financial Industry

In the process of studying the financial industry, Joshua Xiang divided the companies in the industry into three categories:

  • “Traditional” financial companies. Including banks, securities dealers, insurance companies and other financial institutions.
  • Technology and finance company, TechFin company. In essence, it is a financial company, but technology has a strong attribute and plays a central role. Such companies include Ant Financial, Jingdong Finance, Suning Finance and Yixin.
  • Another category is financial technology companies, namely FinTech, such as solution provider tongdun, which do not provide financial products but provide technical solutions for the financial industry.

In the era of digitalization and intelligence, the financial industry will inevitably face the demand of digitalization transformation, among which the demands of traditional financial companies are more intense. The challenges faced by the financial industry, especially traditional financial institutions, in the process of transformation mainly include the following aspects:

  • The way of thinking needs to be changed. Traditional financial institutions tend to be conservative in technological innovation due to the huge volume of data and the specific principle of safety first in the financial industry.
  • Lack of skilled personnel. Artificial intelligence and data science are in the ascendant. Relevant technologies are still in the stage of development and perfection. The corresponding AI and big data talents are still in short supply.
  • It is difficult to access new technologies. Traditional financial institutions have a long history of development, complicated structure and difficulty in accessing new technologies.

Technology leader Should Lead Transformation with Forward-looking and Decision-making Power

In more than 20 years of work, Joshua Xiang has also accumulated rich leadership experience. On the topic of how technical leaders keep their technology forward-looking, Joshua Xiang believes that the most important thing is to keep learning. After the rise of any technology, the first thing to look at is where it applies and what specific problems it can solve. One misconception that technical people are prone to fall into is to use the latest and coolest tools to solve problems that do not require complex solutions, which should be avoided. Technical leaders should track new technologies in time, sink into specific industries and scenes, select technologies according to scenes, and avoid blindly following the trend.

When asked, “What factors will affect the implementation of technological decisions in the process of technological transformation of enterprises, and how will technological leaders deal with them?” At that time, Joshua Xiang believed that there were three aspects:

  • Clarify technical advantages and avoid type selection disputes: some technical personnel have tech religion (technical religion/belief), and there will be disputes about technology type selection in the process of technological transformation. at this time, technical leaders should combine their businesses to clarify the advantages and disadvantages of new technologies.
  • Top-down promotion and bottom-up realization: Technology transformation needs top-down promotion, and clear goals and reach consensus with technical team members to jointly realize technology transformation.
  • Avoiding the risk of change and balancing the pace of change. Change means going out of the comfort zone, and there may be risks in the transition between new and old technologies. Technical leaders should control the overall situation and grasp the balanc e, not rashly advance or procrastinate, and encourage the team to pass through the painful period of change.

“AI is naturally suitable for finance”

Finance Should Be Combined with AI and Big Data

Joshua Xiang believed that the financial industry was born to deal with numbers and had very high requirements for digitization. The advantage of AI is that it can analyze and calculate massive data, build models and do things that no one can do. At the same time, AI’s lack of emotion is a huge advantage in the financial industry, which needs to be highly rational, non-emotional, and must overcome greed and fear. For example, AI’s decisions are more rational and correct without personal emotional considerations when quantifying investments and reviewing credits. In addition, compared with traditional industries such as manufacturing, the financial industry has a higher degree of informatization and a better digital foundation, making it easier to transform to intelligence.

Yixin itself has the genes and attributes of traditional finance. in order to realize intelligent transformation, Yixin has proposed AI financial strategy according to its three major business sectors-inclusive finance, wealth management and asset management: AI technology is adopted according to the problems encountered in its own business scenarios. For example, by analyzing the lender’s demand and repayment ability to recommend appropriate interest rates and products; Analyze the needs of wealth management customers and tailor the financial products they need. The real-life service of Yixin’s wired stores and the virtual service on APP can be used to build models, build user portraits, analyze requirements and recommend products when providing services to customers. These are all based on massive data.

Yixin Big Data Finance

data collection

Yixin’s business covers both online and offline. Its products are all published online and subscribed online, all of which are digitized. Below the line, every interaction with the customer is retained by voice or text, and the activity is also run through by data. A clear portrait of the user is created according to the user’s online and offline behaviors, thus realizing accurate customer orientation. All the collected data and information are stored in the knowledge base to form a user knowledge map, which provides support and basis for decisions of different business lines.

Data usage

How does Yixin specifically combine big data with its own business? Joshua Xiang introduced that the collected user data are mainly used for the following aspects:

  • Personalized recommendation. For example, the above-mentioned wealth management products are tailored for wealth management users.
  • Intelligent wind control. Yixin has its own set of risk control system, which is used to assess investment risks, user credit, etc. and can accurately identify fraud risks.
  • Intelligent push. Yixin has developed an intelligent customer service robot, which uses a set of speech skills simulating real people to complete such services as message pushing, consultation and reminding.
  • Intelligent care. According to the user’s risk tolerance and investment return expectation, a set of investment portfolio strategy is calculated by using the algorithm model, and automatic warehouse adjustment is tracked and recorded every month.
  • Intelligent investment research. Many customers who invest in the stock market need to know more about the industry and specific companies. Intelligen t Investment Research can present information to customers through collecting and sorting.

Yixin data center

Yixin has its own big data platform and is currently building a data center. According to Joshua Xiang, the data center of Yixin integrates data from different sources for cleaning, sorting and correlation to form a financial knowledge map or financial brain for use by different business lines.

The data center of Yixin is still under construction, and many components of the underlying platform have been completed and are open source. For example: big data streaming platform Wormhole, big data bus DBus, big data virtual miscalculation service platform Moonbox, big data visualization application platform Davinci, etc.

Yixin AI central station

At present, Yixin has put AI technology application in an important strategic position and put forward the slogan of “financial embracing AI”. In the process of implementation, Yixin is also developing an AI middleware besides building a data middleware.

What is AI Zhongtai?

In Joshua Xiang’s view, the aim of China and Taiwan is to provide reuse capability for shared services. No matter in technology or business, it is necessary to turn public parts into reusable services for external supply. These services can be capabilities, algorithms, models or a set of tools. The AI console is to use AI’s capabilities as components, services and tools directly into other external application scenarios.

At present, many enterprises have problems of slow response and low efficiency in the research and development of intelligent applications. However, many AI R&D teams are more like outsourcing teams. They build their own positions according to the needs of different business departments and gradually conquer their goals. The process is repetitive and the efficiency is limited. Therefore, AI China Taiwan aims to provide a powerful AI capability support center, which can provide fire support quickly according to business needs and reach the target quickly.

Yixin’s definition of AI medium station is:

AI Central Station is an intelligent model life cycle management platform and service configuration system. It helps front-office businesses to quickly build personalized intelligent services through sharing and reusing intelligent services and automatic management of service research and development.

The hierarchy of stations in Yixin AI is shown in the following figure:

The AI middle station is located between the data model service and the business solution, connecting the business up and communicating the data down. Each level has its own reusable mechanism. The middle part is divided into three parts from top to bottom: business understanding, model learning and data processing. The operation monitoring on the right side encapsulates the products and models in a unified way and provides a unified access interface to the outside world. On the left is platform management that runs through the entire process, including role permissions, tenant management, process control, resource management, etc.

Yixin’s years of experience in the financial industry have precipitated many technologies. For example, Yixin has its own set of tools, processes and platforms suitable for financial business, and has corresponding algorithm libraries and model libraries. With the establishment of the AI middle platform, the construction and modeling of financial knowledge maps can be completed based on the AI middle platform. Including intelligent customer service, intelligent electricity sales, intelligent collection, wind control model and anti-fraud model are all the application scenarios of the AI central station, which are developed and practiced based on the capabilities of the AI central station. At present, AI Zhongtai has been successfully used in Yixin’s intelligent throwing robot (throwing rice RA) and other products.

In the future, Yixin’s AI platform will be packaged into an open platform for the industry, even open up new sources, and discuss the construction together with its peers after it has been internally verified and matured.

“AI era has come, we must embrace boldly and practice carefully.”

Write at the end

As a battle-hardened technician, Joshua Xiang witnessed technological changes and the process of “internet plus” mode and business switching from offline to online. With years of insight into the industry and a forward-looking grasp of technology trends, Joshua Xiang is very optimistic about the integration of AI technology and the financial industry. In Joshua Xiang’s view, AI era has come. We should take AI very seriously, especially its application in financial science and technology. We should embrace it boldly and practice it carefully.

Introduction by Interviewer

Joshua Xiang, CTO of Yixin and executive director of the Science and Technology Commission, has more than 20 years of professional experience, including 18 years of studying and working in Silicon Valley in the United States. After graduation, Joshua Xiang successively worked in Cisco, CA, Dell and other companies in software and network. Joined Microsoft in 2010 and served as assistant to the chairman of Microsoft Asia Pacific research and development group and the group’s technical strategy director, chief project general manager of cloud computing and enterprise division, focusing on the application and landing of cloud computing, IoT and AI. After returning to China, he entered the field of smart retail, taking the post of CTO before siku network, and then as executive vice president of suning cloud business IT headquarters and president of suning technology research institute. he led the development of suning’s unmanned stores, smart speakers, smart shopping guides and other businesses, and applied computer vision and artificial intelligence natural language and other technologies to retail scenes. In 2018, Joshua Xiang became CTO of Yixin, combining AI with the financial industry to explore and innovate.

Interviewer | Joshua Xiang

The original release | AI Front Line