On December 19th, during the "2024 Artificial Intelligence Big Model Financial Field Demonstration Scene and Innovative Application Case Achievement Special Exhibition" event week of the 6th Shanghai Fintech International Forum, guided by the Shanghai Fintech Industry Alliance Innovation Supervision Joint Laboratory and hosted by Qifu Technology, the themed achievement exhibition "Tackling Fintech Innovation and Big Model Application, Releasing Digital Finance into New Forces" comprehensively showcased Qifu Technology's innovative practices in the field of fintech. At the same time, the Innovation Supervision Laboratory of Shanghai Fintech Industry Alliance and Qifu Technology have reached an innovation partnership cooperation on site. Both sides will jointly explore the path of fintech innovation, promote the efficient and secure application of new technologies in the financial field, and assist in the construction of Shanghai's global fintech center.
Qifu Technology CEO Wu Haisheng stated at the special exhibition that technology is the key to driving financial innovation, breaking traditional boundaries and meeting diversified financial needs with intelligent and personalized solutions. He stated that Qifu Technology will deepen research and development, strengthen cooperation with financial institutions, seize this once-in-a-lifetime opportunity for change, and actively explore the infinite possibilities of financial technology. The financial industry is naturally suitable for the implementation of large models, and Qifu Technology will be brave enough to try and utilize the infinite potential of large models to bring tangible efficiency improvements to the financial industry.
Ju Xiaoyan, Deputy Secretary General of the Shanghai Fintech Industry Alliance, introduced that since its establishment in 2020, the alliance has always adhered to the principles of "open communication, cutting-edge display, win-win cooperation, and ecological innovation", actively building a technology cooperation and industry promotion platform with an international perspective, accelerating the coordination and ecological construction of the fintech industry, and providing necessary support for Shanghai to build a global fintech center. In the future, we hope that Qifu Technology will actively participate in the innovation ecosystem construction of the alliance, seize the great opportunity of digital finance development, and gather new forces and contribute new momentum to the construction of financial technology centers.
Song Rongxin, Chairman of Qifu Technology Committee, elaborated on the company's technological concept of starting from the foundation of security and deeply implementing technology in various scenarios, becoming the core driving force for improving the quality of financial technology services. At the model level, by building an enterprise level knowledge base and AI Flow and AI STUDIO, intelligent agent development can be launched as quickly as one day, achieving matrix based coverage of innovation and intelligent weight shaping business processes. Among them, the JARVIS AI generated code model by financial experts accounts for up to 30%, and the AI generated test cases account for 40%; Qifu ChatBI builds a new paradigm for financial technology BI through NL2SQL, semantic modeling, and other capabilities. Qifu Technology uses a technology lossless cold chain model to achieve hourly end-to-end capability transmission with partner banks, and jointly builds a business ecosystem with the support of the FocusPRO digital inclusive credit system that includes risk control, customer acquisition, credit business, growth and other platform capabilities.
Fei Haojun, Chief Algorithm Scientist of Qifu Technology, deeply analyzed the three elements of the implementation of the financial big model product matrix: firstly, deepening the cultivation of scenarios. Taking Qifu Copilot, an agent efficiency improvement assistant, as an example, it accurately locates business scenarios, deeply analyzes hundreds of millions of historical call records, extracts efficient communication strategies and speech templates, and helps agents more accurately and quickly capture user intentions; The second is the data flywheel. Taking the relationship graph constructed by large models in small and micro finance as an example, through a continuous iterative data feedback mechanism, the model performance is ensured to form a virtuous cycle; The third is multi-agent collaboration. Qifu marketing agent can currently call multiple agents to collaborate, achieving the upgrade of marketing from task optimization to autonomous intelligent problem-solving. The three elements complement each other and together constitute the core competitiveness of Qifu Technology in the field of large models.
During the event, Qifu Technology released its AI product Xiaoqi, an AI companion based on big model technology that can provide 24/7 service, deeply understand and predict users' financial needs. Xiaoqi is trained based on the massive financial transaction data, behavior logs, and customer service dialogue corpus accumulated by Qifu Technology over 8 years. According to Lu Yao, Vice President of Product at Qifu Technology, Xiaoqi has full process intelligence without human intervention, multiple rounds of interaction, and language control functions, which can improve the efficiency of user quota application and marketing, and reduce the entry rate.
As the core platform for Qifu Technology's technology output, Qifu Digital Technology officially released the FocusPRO 2.0 digital inclusive credit solution at this special exhibition, aiming to solve the problems of risk, cost, and scale in the field of small and micro credit. Zhou Xuqiang, President of Qifu Digital Technology, stated that the key to breaking through the predicament is to reconstruct the cost structure and control risks through digital and intelligent technologies. FocusPRO 2.0 not only optimizes consumer loan services, but also expands to small and micro business loans, meeting the more diverse and refined financial needs of the market. From January to September this year, with the assistance of FocusPRO, the monthly compound growth rate of Qifu Shuke's loans reached 14%, indicating strong business growth momentum.
Yang Tao, Deputy Director of the National Laboratory of Finance and Development at the Chinese Academy of Social Sciences, stated that the central bank and seven other departments jointly issued the "Action Plan for Promoting High Quality Development of Digital Finance", which not only mentions "strengthening digital technology support capabilities" to assist in the digital transformation of the financial industry, but also highlights the need to "use digital technology to improve the quality and efficiency of financial services in key areas". We can see that with the rapid evolution and iteration of new technologies, the financial industry, as a special information processing industry, has been greatly impacted and affected in terms of its core elements, basic functions, operating modes, risk management, etc. On the premise of ensuring compliance and security, applying cutting-edge technologies more widely to financial service innovation has become an important "lever" for high-quality financial services to the real economy. Especially with the deepening of financial applications of generative AI models, although there are still many challenges, it can be foreseen that it will bring positive effects on improving the automation and intelligence level of the financial industry, as well as decision-making, management, and risk control efficiency.
Dong Ximiao, the chief researcher of Zhongguancun Internet Finance Research Institute, believes that the Action Plan for Promoting the High Quality Development of Digital Finance puts forward the overall requirements and specific measures for the development of digital finance, which defines the direction and goals for the development of digital finance, will accelerate the digital transformation of financial institutions, and promote the healthy development of digital finance. If financial institutions and technology companies can deeply utilize generative artificial intelligence (GAI) represented by large models, vigorously promote digital transformation, continuously innovate products and services, and continuously improve user experience, they may embark on a differentiated and distinctive development path. In this sense, big models will not only become the digital workforce of the financial industry, but also promote the digital transformation of financial institutions and drive the deep development of digital finance.
Qifu Technology stated that the value of AI big models in the financial industry has gradually upgraded from cost reduction and efficiency improvement to efficient enhancement of core business revenue. This will drive disruptive product and service model innovation across the industry, gradually achieving highly personalized financial services seamlessly embedded in users' lives. In September of this year, Shanghai released the "Action Plan for High Quality Promotion of Global Fintech Center Construction in Shanghai". As a fintech enterprise established in Shanghai, Qifu Technology will continue to increase its research and investment in AI big model technology, utilize its strong capabilities in data processing, natural language understanding, decision optimization, etc., deeply explore the potential value of AI technology in financial business, promote the iteration and upgrading of domestic financial big model technology, contribute to the development of fintech in Shanghai, and help the high-quality development of China's digital finance.