• Products
  • Projects
  • Free software
  • R&D
  • About us
Automation of enterprise business processes and workflow with E1 Evfrat.
With a throughput rate of 30 waybills, 50 invoices, or 100 insurance policies per minute, our robot Cognitive Forms is able to recognize, read, and transmit over 80 types of documents.
From requirements to results: purchase requisitions, sourcing event schedules, tenders, contracts in the modular suite Cognitive Lot.
Integrated solutions for nomenclature systems, directories and classification hierarchies — from statement of objectives to loading of data into your ERP.
Able to read and recognize passports with MRZ and without, driver's licences, images and copies thereof — within only 2 seconds with 99% quality. It works faster with NVIDIA CUBA and can be embedded in any system as a .NET component.
Everything in its place. Documents, power point presentations, video and audio — quick and secure deployment on MS SQL and Oracle in full conformity with the State Standards of archival storage.
Intelligent text recognition system. Program for optical recognition of printed text and its conversion into editable form.

Cognitive Technologies
117312, Russia, Moscow
Pr. 60-letiya Octyabrya, 9, office 709

  • +7 (495) 956-90-06
  • +7 (499) 135-5088 (phone/fax)
  • +7 (499) 135-5510 (phone)
  • +7 (499) 135-8968 (phone)






The project's primary purpose was to develop a system for the entry of large amounts of information into the personalized record-keeping system (‘PRKS’) of the Pension Fund of the Russian Federation in order that there could be a single database containing data on all working citizens that allowed to keep records of their income.


Pension Fund of the Russian Federation (PFR)


Capture and processing of documents

Technologies and products involved

  • CognitiveForms


In Moscow alone, adoption of the automation system enabled processing of over 2 million documents a year in timely fashion.

Pension Fund of the Russian Federation (PFR)

Automation System for Capturing and Processing Multipage Employers’ Questionnaires

About the customer

Pension Fund of the Russian Federation (PFR) is one of the most important social institutions of Russia. It is the largest federal provider of social security in Russia.

About the project

The analysis of all organizational and technological processes undertaken in the course of a preliminary examination attributed the greatest share of the costs and time expenditures involved to the manual entry of information into databases. It was established that 30-50 operators was not enough to enter information by hand during peak hours.

The necessity of handling multipage documents exacerbated the situation.

The work on the project included:

  • development of machine-readable forms of different types of PRKS documents;
  • enablement of batch scanning using a high-capacity industrial scanner;
  • adjustment of the system to the concrete documents then in use (questionnaires being printed at different publishing houses and, therefore, differing in size and print quality);
  • development of software to sort out incoming documents by type and to bundle sets of multipage documents;
  • development of software to process packages of documents rejected at the sorting stage (re-scanning if document forms are stuck together, or returning packages to employers with a description of the error);
  • development of software to automatically identify the type of a document form;
  • design of converters to export data into the unique database formats of Pension Fund divisions;
  • development of software to carry out the statistical analysis of input results.

Fig. Examples of the documents involved

The results achieved

In Moscow City alone, adoption of the automation system enabled processing of over 2 million documents a year in timely fashion. Before adoption of the automation system a manually working operator managed to process 100-120 questionnaires (A4 format) per 8-hour shift. After the introduction of the automation system an operator was able to check the accuracy of 500-600 questionnaires per shift. In time productivity was forecast to reach 800-1,000 questionnaires per shift. Hence, the productivity of document entry rose 4-5 times for each operator. Within the first four months of the system's operation 3 million questionnaires were processed, which would have been impossible before the adoption of the system. The automation system was so designed as to be easily adjustable and scalable.