Group of companies continues to invest in AI: new technology speeds up document processing

Group of companies continues to invest in AI: new technology speeds up document processing
Photo is illustrative in nature. From open sources.

With the introduction of the Biorg platform, the personnel department of Agropromkomplektatsiya Group of Companies does not need to manually transfer information from documents, which eliminates the possibility of data errors. After scanning, the recognized information enters the electronic card of the HR document management system 20% faster. The reliability of the processed data is above 99%. The recruitment time has been reduced to 16 minutes.

The platform allows you to conduct analytics on recognized documents, errors and problems of specific scanning points, and on the volume of documents passing through each point. This gives an understanding of how effectively distributed front offices work.

“We are actively piloting computer vision technologies in the digitalization of production. For example, we want to use them to determine the efficiency of the packing department employees, to monitor the HEALTH of animals on our farms. At the same time, in any automation and digitalization projects, the company focuses on payback, direct or indirect. The introduction of the Biorg platform has become one of the innovative projects in the field of HR digitalization. The solution improves the quality of data, the reliability of reports, the speed of their receipt and, accordingly, the speed of making managerial decisions, which ultimately helps to save money. Taking into account the HR electronic document management that is being introduced now, the efficiency of the process will increase many times over,” notes Alexey Leonov,

Technology has shortened the wait time for candidates to come in for a job, giving them a better impression of the employer.

“Through our platform, we recognize almost any type of data, we can verify it and add value by automatically uploading it to various specialized databases. This makes the solution cross-industry, interesting for both commercial players and government departments, for everyone who is faced with the task of making management decisions and developing territories based on data,” notes Ruslan Aligadzhiev, CEO of Bioorg.

In the project for Agropromkomplektatsiya, the Beorg Smart Vision platform automatically recognizes the type of document (passport, TIN, SNILS, etc.) and its content (name, date of birth, series and number of the passport, address, etc.), as well as transfers data to 1C ZUP (Salary and Personnel). At the next stages, employees can get the necessary information through the corporate chatbot: order certificates, receive pay slips, etc.

The values ​​recognized by the neural network with a low threshold of confidence are manually verified on the crowdsourcing part of the platform - Bioorg's own development is certified by the FSTEC and licensed by the FSB. Thanks to the verification module, it was possible to carry out further training of the system as part of a project with the Group of Companies. According to Alexei Leonov, revision was required in connection with the recognition of diplomas, certificates, certificates and other documents of different years of issue.

“People of different ages and professions are employed in our company - from workers to managers. At the moment, the service can recognize all types of documents that are required when hiring new employees,” notes Alexei Leonov.

Beorg Smart Vision is based on Optical Character Recognition (OCR) technology. It helps to convert into electronic, machine-readable formats with the possibility of further image processing from photographs of printed text, PDF files, scanned documents, to digital text.

This technology was among the first to be used by organizations that work with a large flow of documents - banks, insurance companies and community centers, MFCs.

“An important selection criterion in favor of Biorg was the final quality of the recognized data and the total cost of ownership of the service. We noted that verification works well on the platform. This gives a significant advantage in recognition compared to solutions that do not have verification - both domestic and Western. The verification module allows you to confidently recognize handwritten documents, quickly train the system in order to scale the service for processing different types of documents,” sums up Alexei Leonov, DIRECTOR of the Department of Information and Digital Technologies.

Additional benefits for departmental employees include a simple, intuitive interface that they do not need help from IT support specialists to work with.


The Agropromkomplektatsiya group of companies is one of the leading agro-industrial associations in the Russian Federation with a closed production cycle “from field to counter”. Specializes in agriculture, animal husbandry, production and sale of MEAT and dairy products. In the list of brands: "Near Gorki", "Dmitrogorsky Product", "Sincerely Yours". It ranks 3rd among Russian MILK producers and 6th among the largest pork producers. The production facilities of the Group of Companies operate in the Kursk, Tver and Ryazan regions. The HEAD office is located in Zelenograd, Moscow. In total, about 12,000 employees work at the enterprises and organizations of Agropromkomplektatsiya in the regions of presence.

Biorg is a Russian developer of intelligent systems based on computer vision and machine learning for document recognition, including handwritten forms and drawings, and marking up data sets (datasets). The company's key development for the digital transformation of the state and business, the Beorg Smart Vision cloud platform, combines a neural network recognition module (AI) and a crowdsourcing service for data verification. The company provides services for intellectual processing of documents and their sets, digitization of archives, passports, personnel and accounting documentation, creation of electronic document storage systems. Resident of the Skolkovo Foundation.

Read together with it: