The impact of AIGC in the field of financial intelligent customer service has already been revealed, and digital humans are accelerating the reconstruction of financial services and promoting new technologies and new means of production to become real productive forces.
Xin Finance. Hong Ruoxin, Yi Leiwen.
At the beginning of 2023, the picture in the sci-fi movie "The Wandering Earth 2" made many people feel the power and potential of digital humans.
Relying on AI technology, she can not only complete daily interactive chats with her father, but even turn the tide at critical moments and become a key person to save the world in critical moments.
Digital Life Diagram Yaya (Source: Official Promotional Video of the Movie The Wandering Earth).
The film seems to inadvertently predict the trend of AI development this year - a more natural human-computer interaction model.
The popularity of ChatGPT large models has also brought about a paradigm shift in human-computer interaction - we have shifted from graphical user interfaces to natural language user interfaces, and from single technologies such as voice and vision to multimodal technology applications.
In the second half of the year, large models competed to accelerate the implementation of the industry. As a small incision for the landing of large models, digital humans are popular in e-commerce and develop in thousands of industries. Among them, service and marketing is an important town in the financial industry, and it has become one of the most concentrated areas for digital human launch.
According to the latest report of the China Banking Association, among the 42 institutions participating in the survey, 11 customer service centers and remote banks have realized the application of virtual digital humans, and 5 are under construction.
The accelerated implementation of digital humans in the financial industry is actually quite representative.
From the perspective of application, the financial industry has a natural landing scenario - the whole process of customer service, including marketing, customer service response, customer operation, etc., which makes the large model have a small but more specific entry point in the process of industrial upgrading, which can more smoothly realize the transformation from AI industry to industrial AI, and promote new technologies and new means of production to become new productivity.
At present, the influence of AIGC in the field of financial intelligent customer service has been revealed, and digital humans are accelerating the reconstruction of financial services.
Historically, the banking industry has been focusing on the digitalization of financial services for a long time. Under the wave of large models, the financial industry, which naturally lives on data, is constantly soaring with technology and applications.
The digital transformation in the financial sector has gone through three phases:
The first is the electronic stage of financial business;In the third stage, smart services are greatly developed. From the launch of text dialogue robots, to the continuous improvement of voice interaction functions, to the continuous occupation of the market by multimodal digital humans, with the advantages of large models in generating content, digital humans have become one of the standard configurations in the application of large models in the banking industry.The second is the stage of financial channel networking;
Third, in the stage of deep integration of finance and technology, the core technologies of the digital economy such as artificial intelligence, big data, and cloud computing are reshaping the financial business process and making up for the lack of services in traditional finance.
According to the report of the China Banking Association, in the field of customer service, digital humans have assumed functions such as consulting services, financial managers, care for special groups and other life and entertainment services in different scenarios.
IDC**, by 2025, more than 80% of banks will deploy digital humans, undertake 90% of customer service and financial consulting services, and understand text, language, and business will become the norm.
If we look a little wider, digital humans are showing more possibilities.
For a long time, the customer service department has been one of the largest cost centers in the financial industry, which is also an important reason why the banking industry has continuously promoted the intelligence of the customer service center in the past two years - not only to improve the off-the-counter rate and replacement rate on paper, but to integrate business solutions and processes into the answering service, so as to achieve cost reduction and efficiency increase in the real sense, and serve the growth needs of financial institutions.
JD Finance's customer service team has always been a pioneer in the financial services industry, and took the lead in applying intelligent customer service to the whole process of service as early as 2016.
With the continuous iteration and extension of technology and products, the customer service team is combining leading AI technology with digital humans to promote the customer service center to achieve another leap - from a single service to the integration and upgrading of service + marketing + collaboration + management, creating a new paradigm of financial marketing service integration.
Taking JD Finance APP as an example, its customer service is facing more than 4Service needs of 600 million users. Under the massive and high-frequency customer demand, digital humans have become the glue that balances high growth and good service - assisting customer service teams to achieve more efficient operation management in terms of process, organization and growth.
Throughout the industry, in many offline service scenarios, digital human employees can already interact with customers through voice interaction, provide business consulting, product introduction, scanning code and other services, and can also assist customers in handling transfer and remittance services on self-service machines.
At the same time, as mobile has become the main position for financial institutions to reach and serve customers, digital humans are also turning to provide full-process companion services.
In May this year, the service digital human Blue Chip, jointly launched by JD Cloud Yanxi and JD Finance, integrated multiple entrances such as self-service, live broadcast, and customer service consultation in the JD Finance app.
When customers encounter problems that they don't understand, digital people can provide 1v1 exclusive services to help customers understand financial products and guide business handling.
For example, it takes an average of 30 seconds to 2 minutes for a real customer service to query information and receive consultations, while a digital human can complete the query in less than 1 second, providing users with real-time feedback.
It is worth mentioning that JD Cloud Yanxi and JD Financial Services team pioneered the service model of combining live broadcast digital human and service digital human, integrating self-service, robot service and manual service capabilities into digital human, and widely applied in various live broadcast and service scenarios such as business Q&A, service guidance, ** inquiry, financial recommendation, account inquiry, etc.
JD's financial services team effectively links real people and AI tools, so that digital humans can respond independently in live broadcasts, and when encountering incomprehensible complex problems, human customer service can seamlessly connect and drive one or more digital human brains in real time, realizing the operation of AI + real people.
JD Finance Digital Human Blue Core
At the same time, the digital human also analyzes customer behavior and transaction data through big data to understand customer needs and preferences, provide customers with personalized products and services, and make personalized services for customers possible.
A noteworthy detail is that although a variety of digital humans have emerged in the financial industry, there are few digital humans that can truly solve user problems in a closed loop and provide personalized services.
The reason for this is, on the one hand, interaction is the soul of digital humans. Without an interactive design system for scenes, no matter how good the large models and AI tools are, they cannot be transformed into productivity in the vertical field.
On the other hand, the completion of customer inquiries in a high closed-loop is inseparable from the opening of a complete financial business system, which requires institutions to build a set of systematic solutions from infrastructure to superstructure.
With the advent of the stock era, the integration of services and growth in the financial industry is the general trend of business development.
As the entrance to the financial metaverse, virtual digital humans will bring new opportunities for digital transformation for remote banks in terms of digital services, human resources, and operation management in the future, thereby bringing richer digital experiences to customers. The China Banking Association believes.
As a new entrance to human-computer interaction, digital humans continue to expand in the depth and breadth of application in remote banking, which will help make the immersive interactive experience of remote banking more anthropomorphic, reduce costs and increase efficiency more significantly, and continuously improve the quality and efficiency of digital transformation in the future.
In the financial industry, where the fault tolerance rate is extremely low, any mistake can cause immeasurable risk spillover, and even systemic risk.
Judging from the current market environment, statistics show that China has given birth to 238 large models in 8 months, and the surging market wave is full of mud and sand.
In particular, the development of large model technology has always been accompanied by the problem of hallucinations – what the model generates may not be consistent with the verifiable real world, or even fabricated facts.
Therefore, the security of data, the reliability and availability of the model, and the level of the input-output ratio are all factors that must be considered by the organization.
To this end, the JD financial services team has precipitated a complete set of prescriptions, summarized its past 10 years of operational experience, including financial knowledge, tools, data and service systems, and formed a three-layer knowledge system of product-scenario-problem, with a total of more than 9,000 intention words, entity words, attribute words and 20,000 high-quality Q&A knowledge points.
At present, this system has been repeatedly tested in the operation and service of JD Finance, and it also provides services to the outside world through digital people Blue Core and Blue Spirit.
According to the data, since the official launch of JD Finance's customer service digital human on May 10, it has served more than 10,000 customers every day, with a cumulative user volume of more than 1 million, and the service satisfaction is nearly twice that of the text customer service robotAmong them, the digital human anchor can automatically solve most of the customer consultation questions and answers with excellent financial expertise and interactive skills, with a 24-hour problem resolution rate of 85%+ and a satisfaction rate of 90%+.
Even when the robot response is unsatisfactory and manual intervention is required, the platform can also realize the user's imperceptible human-machine collaboration, avoiding the experience lag and unsmoothness caused by service switching.
Today, customer service is expanding from a single text-to-speech contact to multimodal interactionsFrom a single and manual reach such as an account manager to an automated insight method for network-wide reach.
In order to ensure services in the midst of growth, market participants need to further combine customer panorama to achieve differentiated and customized financial product and service applications, form more efficient integrated solutions, and ultimately promote the customer center to realize the leap from cost center to value center.
It is undeniable that the application of any new technology may bring new risks, and from the perspective of industry development, it is obviously unwise to choke on food, just as OpenAI released "Our Method of Ensuring AI Security" as early as the beginning of the year, responding to questions about the safety of its AI models.
In Son's words, within 10 years, there will be a gap between people who can use AI and those who can't, and people and monkeys. Whether it's in the financial industry or at the broader market level, we will embrace change.