Automation of Verification

Fast and convenient for the client to check his creditworthiness when issuing a bank loan involves the recognition of his personal documents for identification. At the same time, risk reduction is provided by high-quality cross-verification of documents by operators. Automation of this process in order to reduce the cost of the contact center may be obtained through the use of modern methods of image recognition.

Input Data

  • Set of images and scans of personal documents
  • Marking images with text boxes
  • The true value of each field
  • Base named entities of the Russian Federation

Output Data

  • Model recognition of the main elements of the passport
  • Model character recognition for each element of the passport

Main Results

The solution is integrated into the business process of the customer:

  • Reduced staffing costs 10%
  • Improving the quality of risk analysis by 15%
  • Increasing the approval application rate of 30%

Pilot tests were also conducted on the following documents:

  • Passport
  • SNILS
  • Driver’s license
  • Technical means passport

Recognition quality:

Surname Recognition Accuracy 85 %
Gender Recognition Accuracy 88 %
Recognition accuracy of numbers and series 73 %
Recognition accuracy of the Authority and date of issue 80 %

Example of operation with recognition on the Russian passport: