ZRobot: Derive Real Value from Data Technology
Yang Qiao, CEOEvery enterprise today claims to be data-driven. Data technology as the foundation of the decision-making process, has evolved as a competitive edge for organizations, and financial enterprises are no exception. However, Yang Qiao, CEO, Zero Machine Technology (ZRobot), begs to differ. According to Qiao, data technology alone (such as machine learning and deep learning, among others) cannot bring in competitive advantage. “It has to be the actual application of data technology creating real value for the financial industry,” says Qiao. Having amassed years of experience in strategy development, modeling, data analytics, financial analysis, operation analytics and reporting, Qiao leads his team at ZRobot—a wholly-owned subsidiary of JD Digits— to navigate the financial industry through digital technology. ZRobot was born with a focus on assisting organizations in realizing the potential of the Internet, digitalization and intelligence, reducing costs, improving efficiency, user experience, and upgrading models.
Using deep learning and machine learning technologies, ZRobot is committed to building a big data ecosystem that focuses on the actual application of data technology. In creating this ecosystem, ZRobot collaborates with over 400 companies in the banking, insurance, securities, fund, consumer finance, financial leasing, auto finance, OTAs, and education sector. Backed by powerful classified modular data resources and integration capabilities, ZRobot equips this ecosystem with services that have been instrumental in lowering operation costs, improving underwriting and operation efficiencies (such as higher approval rates), and reducing default rates. From risk control to anti-fraud, the company covers diverse aspects yet remains singularly focused on the application of data technology and corresponding results in the financial industry.
In a bid to power the application of data technology at financial enterprises, ZRobot brings in a full suite of services from standard risk scores, SaaS platforms for risk management and BI platforms to customized risk policy and model building consultation. The company has made significant strides, providing the clients withits flagship products that include credit scores and customer profile labels like ZR Risk Score, ZR Consumer Lending Risk Score, Financial Profile labels, Online Shopping Profile Labels, and more.
The risk scores help financial institutions make better underwriting and portfolio management judgment and the profile labels assist them to further finetune their risk strategies to improve their risk performance. Based on advanced machine learning technology, ZRobot has conducted self-learning and integrated a billion-level anti-fraud database to develop its intelligent anti-fraud platform.
Added to that, the company also strengthens anti-fraud initiatives in enterprises through anti-fraud scores and profile labels that include the ZR Fraud Score, ZR-patented Zebra Spread Score, the blacklist, over-indebted indicators, address stability index, account stability index, among others.From the risk control and marketing perspective, ZRobot has developed marketing products and various SaaS platforms for risk and anti-fraud decisions, and customized modeling, data interface, data visualization and credit management. All these solutions and services from ZRobot are layered with consulting services for product design, risk control process and risk policy/strategy and model building.
Keeping Abreast of Developments
While the government in China has increased emphasis on data security and released corresponding laws and regulations to regulate the use of individual data, ZRobot follows the regulations closely and ensures compliance. With products and services based on data acquired with individual customer’s direct authorization, ZRobot complies to the regulations with the aid of its compliance teams and JD Digit’s (JD.com’s technology arm) internal control and compliance teams separately. Therefore, the usage of data and the output of data products are guaranteed to meet the government’s requirements and standards.
ZRobot has been working closely with Baihang Credit—its biggest competitor in the marketplace—andhas turned this competitive relationship into an alliance. ZRobot has reached an agreement to couple its products and services with those of Baihang Credit. This collaboration is also highly beneficial for Baihang Credit since ZRobot has access to data of the world’s second-largest e-commerce platform JD.com.
The partnership has enabled ZRobot to provide better and more comprehensive products and services primarily for top financial institutions, such as ICBC and ABC.
While the government has ordered P2P platforms to either shut down or transform into consumer lending companies, ZRobot changed its focus to serve national banks, regional banks, joint-stock banks, licensed consumer finance companies, and institutions instead of small and unlicensed clients such as internet finance companies and P2P platforms. ZRobot is committed to utilizing massive high-dimensional data resources, combined with advanced data mining technology and model algorithms in the industry to provide partners with intelligent risk management solutions to enhance the overall risk control capabilities of the company. By combining JD.com and JD Digits’ consumer authorizedata such as transaction and behavior data, ZRobot can cover majority of the population that is not currently covered by the Central Credit Bureau since there are over 600 million registered users on JD.com and JD Finance (which is the finance branch under JD Digits) and over 340 million active users. By applying advanced data technology such as machine learning and deep learning, ZRobot can evaluate the applicants’ potential credit risk and fraud risk especially for those that are not covered by data in the Central Credit Bureau. By working with various financial institutions, ZRobot can help these people get fair credit without putting additional risk on the lending financial institutions. Moreover, ZRobot’s data technology covers all lending processes from marketing and customer acquisition to collection. As a result, with ZRobot, financial institutions can enhance the overall risk control capabilities and reduce operating costs.
Of the Past and the Future
In a recent client engagement, ZRobot helped one of China’s largest auto finance companies design and build a customized auto finance decision engine. Before the decision engine was put into place, each application took two business days to process on average and needed to involve human employees from seven different departments. This client engagement was a resounding success wherein after the decision engine went online, operation cost was reduced by over 70 percent, and the whole process was shortened to within half a day.
Another example: ZRobot helped one of China’s largest regional banks to build its decision engine to cater to its online lending business since the regional bank used to use traditional risk management measures such as human review of applicants’ employment verification, bank statements, etc. and its business focuses on offline products. At the same time, ZRobot worked with the bank’s risk management team to build customized risk strategies and models so as better assist online consumer lending risk control. From the time the online product went live to the end of 2019 (less than 1 year), the online portfolio has grown to 10billion RMB with default rate less than 0.5%. The bank has presented our case to the People’s Bank of China (China’s central bank) as a success cooperation story between a technology company and a traditional bank.
Such customer success stories only validate the efficiency of ZRobot—a company that has only 142 employees yet can compete with the biggest competitors that have over 1000 employees. Besides, ZRobot’s businesses span B2B, B2C and B2B2C which are far wider than its competitors. The uniqueness of ZRobot stems from its strong technical team that brings in diverse experiences earned from Fortune 500 companies, internet companies, consulting companies, and the like. Added to that, the team adopts unique data mining and modeling techniques and holds deep understanding of different industries. This helps ZRobot to work with partners from different industries, not just financial industries and has formed an eco-system with all the cooperating companies. ZRobot’s ability to combine technology with different application scenarios (both internal and external scenarios)is a noteworthy differentiator—a skill that brings in technical improvements and break-throughs all come from practice and not theory. ZRobot’s distinctiveness also lies in the huge data resources it utilizes to build up its data technology. JD Digits as ZRobot’s controlling shareholder, generates 800 TB of data daily. The quantity is larger than what a small financial company can gather in a year. ZRobot is positioned to work with partners from different industries, not just financial industries and has formed an eco-system with all the cooperating companies.
In the next two years, while the authorized user base in B2C businesses will reach 400 million, Qiao expects this user base with their direct authorization, will have a credit score, which will make Xiaobai credit a commonly accepted and unified social credit score in China. At the same time, ZRobot’s new consumer product leasing platform, Jingxiaozu, will become China’s largest consumer product leasing platform with annual orders over 300,000 and GMV over $300 million. Besides expanding its businesses to South-east Asia and South Asia, Eastern Europe, and South America, the company also looks forward to becoming one of the top three data technology and fintech service providers in China. In terms of the product roadmap, ZRobot will continue to improve its current flagship products.
While the government has ordered P2P platforms to either shut down or transform into consumer lending companies, ZRobot changed its focus to serve national banks, regional banks, joint-stock banks, licensed consumer finance companies, and institutions instead of small and unlicensed clients such as internet finance companies and P2P platforms. ZRobot is committed to utilizing massive high-dimensional data resources, combined with advanced data mining technology and model algorithms in the industry to provide partners with intelligent risk management solutions to enhance the overall risk control capabilities of the company. By combining JD.com and JD Digits’ consumer authorizedata such as transaction and behavior data, ZRobot can cover majority of the population that is not currently covered by the Central Credit Bureau since there are over 600 million registered users on JD.com and JD Finance (which is the finance branch under JD Digits) and over 340 million active users. By applying advanced data technology such as machine learning and deep learning, ZRobot can evaluate the applicants’ potential credit risk and fraud risk especially for those that are not covered by data in the Central Credit Bureau. By working with various financial institutions, ZRobot can help these people get fair credit without putting additional risk on the lending financial institutions. Moreover, ZRobot’s data technology covers all lending processes from marketing and customer acquisition to collection. As a result, with ZRobot, financial institutions can enhance the overall risk control capabilities and reduce operating costs.
Of the Past and the Future
In a recent client engagement, ZRobot helped one of China’s largest auto finance companies design and build a customized auto finance decision engine. Before the decision engine was put into place, each application took two business days to process on average and needed to involve human employees from seven different departments. This client engagement was a resounding success wherein after the decision engine went online, operation cost was reduced by over 70 percent, and the whole process was shortened to within half a day.
Another example: ZRobot helped one of China’s largest regional banks to build its decision engine to cater to its online lending business since the regional bank used to use traditional risk management measures such as human review of applicants’ employment verification, bank statements, etc. and its business focuses on offline products. At the same time, ZRobot worked with the bank’s risk management team to build customized risk strategies and models so as better assist online consumer lending risk control. From the time the online product went live to the end of 2019 (less than 1 year), the online portfolio has grown to 10billion RMB with default rate less than 0.5%. The bank has presented our case to the People’s Bank of China (China’s central bank) as a success cooperation story between a technology company and a traditional bank.
Such customer success stories only validate the efficiency of ZRobot—a company that has only 142 employees yet can compete with the biggest competitors that have over 1000 employees. Besides, ZRobot’s businesses span B2B, B2C and B2B2C which are far wider than its competitors. The uniqueness of ZRobot stems from its strong technical team that brings in diverse experiences earned from Fortune 500 companies, internet companies, consulting companies, and the like. Added to that, the team adopts unique data mining and modeling techniques and holds deep understanding of different industries. This helps ZRobot to work with partners from different industries, not just financial industries and has formed an eco-system with all the cooperating companies. ZRobot’s ability to combine technology with different application scenarios (both internal and external scenarios)is a noteworthy differentiator—a skill that brings in technical improvements and break-throughs all come from practice and not theory. ZRobot’s distinctiveness also lies in the huge data resources it utilizes to build up its data technology. JD Digits as ZRobot’s controlling shareholder, generates 800 TB of data daily. The quantity is larger than what a small financial company can gather in a year. ZRobot is positioned to work with partners from different industries, not just financial industries and has formed an eco-system with all the cooperating companies.
In the next two years, while the authorized user base in B2C businesses will reach 400 million, Qiao expects this user base with their direct authorization, will have a credit score, which will make Xiaobai credit a commonly accepted and unified social credit score in China. At the same time, ZRobot’s new consumer product leasing platform, Jingxiaozu, will become China’s largest consumer product leasing platform with annual orders over 300,000 and GMV over $300 million. Besides expanding its businesses to South-east Asia and South Asia, Eastern Europe, and South America, the company also looks forward to becoming one of the top three data technology and fintech service providers in China. In terms of the product roadmap, ZRobot will continue to improve its current flagship products.
