Size daha iyi bir deneyim sunmak için çerezleri kullanıyoruz. Ziyaretinizi varsayılan ayarlarla gerçekleştirerek, çerez kullanımına izin vermiş oluyorsunuz. Çerez politikasını inceleyin.
Xianhui Li2019-08-21 94
Data, algorithms, scenarios — all are continuously interacting, in turn integrating more closely and driving new developments in Artificial Intelligence (AI). Simulating human behaviors and outcomes, AI is now capable of rapid processing and self-learning.
With the development of the technology, AI applications have not only penetrated the financial services industry — they have matured. In doing so, they are proving to be disruptive forces across banking and insurance, as well as capital markets.
For example, surveys suggest that approximately 23% of jobs in the financial sector in China — where AI financial practices are active — will be replaced by AI. Furthermore, of the remaining 77% of jobs, working hours will reduce by approximately 27% due to the influence of AI. That figure is equivalent to an astonishing 38% increase in efficiency.
Perception technologies such as computer vision, voice recognition, and Natural Language Processing (NLP) are maturing and being applied to financial service processes. As a result, service processing automation in the industry has reached new levels, in turn improving customer experiences.
Typical examples of AI-powered automation include chatbots, automatic identity authentication, and intelligent bill entry via Optical Character Recognition (OCR).
A chatbot follows the standard path of the customer journey, analyzes conversations, and understands the intentions of these conversations using machine learning algorithms. When encountering difficulties, the chatbot sends any issues to live personnel to resolve, learning from the manual replies that are sent back in return —all improving the quality of customer service as well as reducing costs.
Automatic identity authentication verifies a customer's identity by analyzing voice, eyes, or facial features. Compared with the traditional approaches of security questions and passwords, this solution delivers a higher verification efficiency and a superb user experience — busy users are not required to remember trivial details.
An intelligent bill entry solution greatly improves error control and entry efficiency, compared with manual record keeping.
Advancing data analysis and deep learning technologies greatly improve the accuracy of intelligent analysis and decision-making, creating or improving the business value of financial products and services.
In the past, business intelligence and traditional analysis only covered trend analysis, cause analysis, data mining, and prediction. Now, AI can enhance the relevance and specificity of suggestions through continuous learning and improvement, enabling personalized analysis. Targeted analysis and decision-making are available in the fields of risk management, marketing, and services. For instance, AI offers credit ratings based on social networking, optimizing existing rating mechanisms and becoming capable of generating scores, even for people without credit records.
AI also supports personalized marketing based on individual customer preferences and specific product DNA, recommending only the most suitable products. In addition, AI is capable of dynamic fraud detection, detecting frauds in real-time from complex transaction modes.
In their marketing efforts, it is vital for financial enterprises to precisely identify real customer needs. AI is a perfect fit for this task, using user profiles and big data models to enable precision marketing. Chatbots can also discover potential customer needs, improving sales conversion rates, customer service efficiency, and the overall user experience — all while significantly lowering labor costs.
Investment institutions and investment banking departments spend most of their time on trivial tasks, from data collection and analysis to report writing. AI systems offer unique advantages in these areas, since they are capable of processing massive volumes of data at speed. The technology is able to use NLP to uncover the hidden patterns governing market changes — for example, which companies' share prices will be affected by the launch of HarmonyOS, Huawei’s latest Operating System (OS).
An AI-powered robo-advisor considers the unique risk preferences and financial status of investors, harnesses big data and quantitative models (in particular, portfolio theory), then provides each investor with individualized asset allocation strategies and wealth management services centering on index funds. The solution also supports position tracking and dynamic adjustment, as the market changes.
Investment advice generated by AI combines investor preferences with Modern Portfolio Theory (MPT), offering transparent information with low commission fees. The solution makes private banking services both accessible and intelligent, allowing ordinary investors to enjoy a level of service once reserved for the high-end.
Intelligent risk control, such as transaction fraud prevention, is also empowered by AI. By learning user behavior patterns and analyzing fraud blacklists, the anti-fraud model can identify any abnormal behavior of special users during loan application and approval processes, effectively detecting fraud. In addition, AI systems are able to obtain knowledge and learn rules from massive amounts of transaction data, to detect and block exceptions such as malicious cash-out, card theft, spam registration, other illicit activities (in marketing), and fake transactions. A secure and reliable financial environment emerges as a result.
Know Your Customer (KYC) is the foundation of all financial services. Voice, fingerprint, facial, and iris recognition technologies drastically reduce both the time to recognition as well as error rates. Moreover, AI systems can draw 360-degree customer profiles using massive volumes of historical transaction data, allied to collaborating external data, helping to improve business risk control capabilities.
As a leading ICT technology provider, Huawei believes that sufficient data, powerful computing power, and quality algorithms lie at the foundation of AI. Based on these building blocks, Huawei is launching a full-stack, all-scenario AI platform solution dedicated to the financial services sector. Together with AI applications, this solution will significantly improve financial transaction service experiences, transaction efficiency, and risk prevention capabilities.
1. Huawei provides Ascend series AI chips with extremely high computing power (up to 256 PFLOPS for a single core) to help financial enterprises build a powerful Atlas server computing platform (up to 1024-core cluster networking). What's more, the platform also comes with a chipset operator library and a highly automated operator development toolkit, to optimize the AI inference framework and computing capabilities, using chip core technologies. The platform is compatible with major inference platforms in the industry.
2. Adhering to the openness principle, Huawei helps financial enterprises build FusionInsight ModelArts, an AI platform based on MindSpore (an efficient deep learning-based inference framework) and ModelArts (an AI application development platform). ModelArts contains universal inference capability development with an invoking interface, enhanced inference capability development with an invoking interface, and a pre-integrated AI inference capability set. Thanks to these modules, the platform maximizes AI application development efficiency and simplifies development processes. Huawei ModelArts is also fully compatible with other inference frameworks in the industry.
3. Data plays a vital role in smart finance. To support AI's data management, storage, and processing, Huawei provides the FusionInsight HD big data platform and the GaussDB200 distributed data warehouse platform.
In addition, Huawei works with ecosystem partners to provide AI modules for financial services based on Huawei-developed AI applications. The regulation-compliant modules include biometric recognition capability, bill recognition capability, financial semantic library, financial semantic emotion knowledge base, financial transaction risk model library, and financial credit rating model library modules.
Huawei provides full-stack AI solutions for large financial enterprises, helping them build a controllable AI platform and develop differentiated financial services and product capabilities.
Using Huawei's solutions, banks can build a a next -generation, intelligent risk control platform to monitor transactions in real time and block abnormal transactions: one Chinese Bank prevented 83% of risky transactions and avoided the loss of more than CNY 100 million (over USD 14 million) within half a year.
For small- and medium-sized financial institutions, AI services are available on AI platforms built by Huawei and industry associations or central banks, helping them to develop their own AI businesses.
The Huawei AI cloud service platform provides more than 50 services and 142 functions, with the aim of developing an inclusive AI platform with the fastest growth possibilities.
Looking elsewhere, to the intelligent policy entry field, YSX helped insurance companies automatically recognize policies, using HUAWEI CLOUD's bill recognition service (via OCR): processing time fell from dozens of minutes to just a few seconds.
Intelligent policy entry, intelligent damage assessment, anti-money laundering, transaction fraud detection, facial recognition payment, chatbot, and robo-advisor, along with other services — all have been successfully supported by Huawei’s AI platform, helping to build the smart finance of the future.
For more information, please visit: https://e.huawei.com/en/solutions/industries/smart-finance