Tencent Financial Cloud Applies “Arithmetic Approach” to Fintech Operations


The head of Tencent’s cloud finance vehicle has touted the use of an “addition, subtraction, multiplication and division” model when it comes to the online giant’s fintech operations.

Hu Liming (胡利明), general manager of Tencent Financial Cloud (腾讯金融云), spoke about the model at the Lendit Fintech Summit held in the Pudong district of Shanghai on 6 September 2018.

Tencent Cloud Financial is currently applying the model to its approximately 6,000 financial clients and more than 150 cooperating banks, including China Construction Bank, Huaxia Bank and the Shanghai Pudong Development Bank.

According to Leiphone 90% of China’s newly established fintech platforms in the past three years have been built using Tencent Financial Cloud, including consumer finance companies, online loan platforms and peer-to-peer lenders

According to Hu the “addition” part of the model consists of using targeted sales to help financial institutions to obtain and retain customers and expand the scope of their operations.

Tencent is well positioned to achieve this, given that it currently has over 1 billion users as well as user data amassed from more than 2 billion different devices.

The company has developed its own sales algorithms that uncovers and supplies links to target demographics based on the tags, images as well as geographic location of users.

During cooperation with Pingan Group Tencent’s targeted sales techniques helped to achieve an 84% increase in click through rates and a 40% increase in  conversion rates.

The “subtraction” part of the model involves reducing the risk of financial institutions via the establishment of online risk control systems, such as its Tencent’s Tianyu AntiFraud system (天御反欺诈).

Hu said that Tianfu provides its services to 300 companies, averting online losses of more than USD$25 billion, while the micro-loan error rate of online lender WeBank, in which Tencent is a key shareholders, is just one percent.

The “multiplication”part of the model involves using a machine-learning platform to provide various AI capabilities, including voice recognition, text processing and machine vision.

These technologies have permitted the development of automated sales and customer service devices, as well as smart investment advisory services.

According to Hu Tencent has helped China CITIC Bank to create a “voice bank” that supports Mandarin Chinese, English and Cantonese, while WeBank makes extensive use of AI for customer service purposes.

The “division” part of the model refers to the use of a cloud computing’s distributed framework to improve IT support needs.

According to Hu over a dozen banks are already using a distributed services framework with a TDSQL database at its core to service more than 20 billion accounts.