Home/Case/ Intelligent Review System for a National Joint-Stock Commercial Bank
Intelligent Review System for a National Joint-Stock Commercial Bank
Requirements
Currently, the document review system is rarely deployed in the banking and financial industries in China. This is because the system requires a wide range of knowledge of document review, its rules vary to a certain extent in the business forms among different customers and across different regions, and it relies heavily on the human experience to sort out the rules. In terms of technical capabilities that support the implementation of business scenarios, the system requires advanced AI technologies.
Solution
Through sorting and classifying professional document-review rules by experts, we summed up nearly 300 rules. OCR technology was employed to recognize pictures and images, NLP technology was employed to intelligently extract message clauses, knowledge graph was used to build a knowledge system network, and a rules engine was used to configure, analyze and execute the rules. These efforts finally led to successful implementation of the document review system in China.
Achievements
The bank successfully launched the document review system in China, which increased recognition efficiency by 100 times and review efficiency by 300 times, greatly improving the efficiency in reviews as well as customer experience, while saving operating costs and updating the bank's IT technology system.