Banks need to reposition themselves and find suitable development strategies to revolutionize their business with the help of FinTech. Because the financial/banking industry tends to be very traditional, however, leading technology companies should focus on the integration of technology innovation and scenario application to achieve success in the finance sector.
Artificial intelligence had a massive impact on the banking industry in 2017. From a technical point of view, artificial intelligence applications can be divided into two specific categories: basic AI and industry AI. Basic AI can be integrated into application systems, like facial recognition, speech recognition, etc. Industry AI has more applications in business, such as anti-fraud, Robo-advice, and so on. At present, mainstream artificial intelligence technology is data-driven machine intelligence. The difference between the two categories mainly lies about who takes control of the data or who uses the data to produce the AI models.
Banks revolutionize their products and processes and replace repetitive work with AI. Production efficiency can be consistently improved (e.g., through the use of smart contracts and Robo-advice). In addition, the introduction of basic artificial intelligence applications such as biometrics in mobile banking, smart counters, and other scenarios can solve the key problem of customer verification. AI not only can optimize business processes but also can greatly enhance the user experience and pave the way for the next stage of development.
Big data intelligence is the next stage of AI development. Technological innovation will bring more use cases, which in turn are supported and driven by big data. In fact, the research and application of big data in banks started before the application of artificial intelligence. The current emphasis is on the integration of basic AI and industry AI so as to provide better services for customers. The combination of big data and basic AI can improve the intelligence level of system products and business processes. Key technologies, however, need to develop independently, including customer profiling, product profiling, behavior analysis, personalized recommendation engines, etc. Banks need to form their own AI R&D capability, which is also the key to using big data in their core competencies.
Furthermore, Artificial intelligence development is to fulfill all-channel intelligent decision making, seamlessly connecting customer identification, behavior prediction, and all other channels and updating dynamic optimization based on customer response. The bank should develop internal consensus and establish an effective cooperation mechanism from business process to system development, from product design to marketing support, and from data analysis to data mining.
In recent years, big data have been widely used in many fields of banking, progressing from data analysis reports to data mining models and then to data products.
Data, use cases, and modeling are the three fundamental elements for banks in applying big data. In fact, it is possible to start data-led product development in any of the three ways. For example, transforming the business into one that uses big data applications might begin with the data analysis used in the bank’s traditional business. Finding the application direction or use from the internal and external data in the bank might relate to big data risk management and marketing. Finally, businesses need innovative models and technologies to solve new problems. According to the 80/20 rule, most big data applications must be derived from business analysis and do not necessarily require “huge” data and “esoteric” technology, which is a major concern in practice.
The widespread use of big data in banks is accompanied by strong, two-part demand for data asset management: 1) the sharing and application of the internal and external data, and 2) data mining, intelligent model management, and knowledge transfer. The first part lays the foundation for the big data application and provides the raw data material; the second part represents the derivative asset, which provides the added value of the big data but also holds the potential to be exported. The management mechanism of these two parts is also a relatively new topic in the industry and is still in its infancy.
Banks that achieve breakthroughs in multiple domains can try to set larger goals (e.g., establishing an efficient data value chain in the field of risk management marketing and other fields) and gradually establish the initial AI framework so as to connect the needs of many fields. A pragmatic AI-brain prototype needs to include customer profiles, product profiles, data mining models, and decision engines. Data mining models are the core of this kind of intelligence. Customer profiles and product profiles provide continuous input to the modeling process. The decision engine transforms the model output into actual business actions. The big data process capabilities are primarily reflected in three aspects: improved customer perception, more intelligent algorithms, and faster decision support.
Cloud computing is the use of the internet to access applications, data or services that are stored or run on a remote server. Typically, cloud computing exists at one of three levels: Infrastructure-as-a-Service (IaaS) at the bottom, Platform-as-a-Service (PaaS) in the middle, and Software-as-a-Service (SaaS) at the top. The infrastructure layer aims at releasing the productivity of underlying resources while PaaS serves as the core platform hosting applications.
The development of the infrastructure layer is constrained by many objective conditions in the bank, including technical barriers, computer facilities, electricity, and other problems. In business and application expansion planning, consideration of how to rectify these problems should happen first. Considering the platform layer of bank cloud computing from the FinTech perspective, besides the standard development middleware, we try to make it the platform of the other three crucial technologies (ABD). Based on the in-depth development of open-source technology, it is possible for the bank to build a platform with independent intellectual property rights. For small and medium-sized financial institutions, it is more economical to use the PaaS platform provided by large institutions than to pour money into innovative applications.
Finally, the software layer directly provides the application corresponding to the business scenarios. It can be cloud banking services, such as cloud payments, or cloud FinTech services, like risk management, marketing, operations, and other intelligent data products, or even more basic data services.
Two factors are limiting the adoption of blockchain:
- The technology is not mature. Its performance, privacy issues, operation, and maintenance are not up to standards for enterprise use.
- The business model is not ready. In a multi-center scenario, it is difficult for different parties to reach consensus.
To be sure, this doesn’t mean blockchain is not good enough, but rather, that good projects are lacking in the blockchain space. Blockchain 1.0 is the digital currency represented by bitcoin; blockchain 2.0 is the smart contract platform represented by Ethereum. Blockchain 3.0 is moving forward in the fields of cryptography, consensus algorithms, cross-chain fusion, performance optimization, and so on.
In addition, the development of fiat digital currency may become a breakthrough point for blockchain. Although there is no direct connection between them, it can be expected that the emergence of fiat digital currency will provide great potential for the development of existing blockchain applications. Finally, compared with AI, big data, and cloud computing, blockchain applications are entirely technology-driven. Adopting such “pure” technology is well worth our effort.