How data science is being implemented in Russia: from banking to mining

The global data market is growing with the amount of information that millions generate daily

users. In 2018, users and businesses created 33 zettabytes (1,021) of information, up to 175 ZB by 2025. According to an IDC report, the global data science market reached $ 166 billion in 2018, and analysts predict it will grow to $ 260 billion by 2022.

Now Russia accounts for less than 2.29% ofof the global market for data science solutions. For comparison, the USA accounts for 35% of its volume. As of 2020, various experts estimate the volume of the Russian market at 10–65 billion rubles. By 2024, according to the forecast of the Association of Big Data Market Participants (ADB), this figure will reach 300 billion rubles.

The main factors holding back developmentRussian big data market, analysts say the high cost of solutions, the lack of quick results from implementation, the lack of specialists, as well as problems with the quality and confidentiality of data. The presence of these barriers leads to uneven development of the market for such solutions. Most often they are implemented by banks, telecom, retail, industry and the public sector. Other industries are lagging far behind.

Let's consider several leading industries in the implementation of data science solutions, as well as industries that have great potential for using tools based on data analysis.

Who is already using data science

  • Banks. In finance, data science solutionshelp, for example, to evaluate borrowers (credit scoring), verify users and prevent fraudulent transactions; analyze the performance of borrowers' portfolios and predict the demand for cash at ATMs.

In the banking sector, most companies createown data science development departments, rather than ordering them from third-party vendors. Products based on data analysis are used by VTB, Sberbank, Tinkoff Bank, Raiffeisen, Uralsib, OTP Bank, Leto Bank and others.

For example, Sberbank has created tools forsegmentation and assessment of customer creditworthiness, cybersecurity, personnel management, forecasting queues in branches and calculating bonuses for employees. Tinkoff Bank has developed a solution for risk management and assessment of the needs of existing and future borrowers. In addition, the bank has introduced a voice assistant "Oleg", which can transfer money at the user's request, order and send electronic documents by mail, as well as book a table in a restaurant and buy movie tickets.

  • Telecom. The collection and analysis of information enables companies fromindustry to predict subscriber churn and find ways to retain them, create personalized offers and determine future network load based on subscriber movements.

The largest operators, like banks, haveown teams of data science solutions development. Such tools are already used by the operators of the “big three”: MTS, Beeline, and MegaFon. For example, MTS has created a system that generates personalized offers based on data on traffic consumption, devices used and purchases of a subscriber. And MegaFon uses data science to improve the quality of services and optimize costs. In addition, the operator has a subsidiary that develops data science solutions for business.

  • The property. In the construction, commissioning and purchase industryreal estate traditionally accumulates a large amount of documentation: contracts, applications, building projects and rights to land. Data science solutions in this area can be used to accurately predict the demand for objects, assess risks when choosing a site for construction and predetermine customer behavior.

Such solutions were implemented by CIAN companies,ROSEKO and DomClick. The latter used a machine learning-based system to speed up inbound processing and improve conversions. Agroholding Miratorg has implemented a solution to automate the process of acquiring real estate for warehouses and farms.

  • Insurance. In the field of insurance, the frequency of interaction withclients and the speed of settlement of issues is much lower than in banking. However, the data accumulated by insurance companies is sufficient to implement technological solutions. For example, data science is used to calculate CASCO and OSAGO tariffs for an individual client, as well as to form a list of services for medical insurance policies.

Among the industry representatives, solutions based on this technology were implemented by the companies AlfaStrakhovanie, RESO-Garantia, SOGAZ, Tinkoff Insurance, Renaissance Insurance.

For example, SOGAZ uses machine learning toidentifying cases of unjustified overstatement of the cost of medical care provided. Automation has eased the burden on medical experts and has allowed the company to cut costs.

  • Jurisprudence. Technologies in this area help automateprocesses of processing documents, checking counterparties, drafting contracts and comparing versions, selecting court practice in the case, as well as providing citizens with more complete access to legal services.

Pravo companies use solutions in this, BrandMonitor, "Amuleks", the Agency of judicial collection, as well as the department of legal support and legal risk management of Business Car Group and others.

For example, the Pravo portal development has created a system that processes typical user requests without the direct involvement of lawyers and independently provides legal support. BrandMonitor has developed a solution for evaluating counterfeit products on the Internet by logo, description and trademark. The system detects the violation and automatically notifies the hosting provider about it, which has a registered website for the sale of counterfeit products. And the Embedika company is developing an intelligent system for checking labor contracts, which, based on the methods of active learning and online learning, is retrained in the process.

  • Logistics. This industry accounts for 5% of world GDP ormore than $ 4 trillion in monetary terms. Along with an annual growth of 7-10%, the volume of data is also increasing: if in 2017 this figure was 16.2 ZB, then in 2020 it will be about 44 ZB.

Data-driven solutions in this areawill help you plan routes more efficiently, predict the profitability of transportation, as well as predict accidents due to wear and tear of equipment and ensure the safety of closed facilities.

Airlines have implemented similar solutions in RussiaAeroflot, S7 and Utair. S7 uses a passenger face recognition system at Domodedovo Airport and a system for personalizing services for visitors to business lounges, while Aeroflot uses tools to segment customers based on their purchasing power to predict future purchases.

  • Public sector. Government structures accumulate the largest amount of data - their processing and analysis can increase the efficiency of work and the quality of services rendered several times.

Data science products implementedFederal Tax Service (FTS), Analytical Center of the Russian Government, Pension Fund, Moscow Government, Mandatory Medical Insurance Fund, Federal Security Service, Investigative Committee and Foreign Intelligence Service.

FTS uses data science solutions for controlfor the payment of value added tax (VAT), and the Moscow government - within the framework of the project to improve the urban transport system "General Plan". In it, models analyze passenger traffic in surface transport and metro. The Ministry of Telecom and Mass Communications is developing a "Digital Platform for Control and Supervision Activities", in which data science solutions optimize the rule-making process.

  • Extractive industry... Big data analysis in this industry allowspartially automate the assessment of the efficiency of field development, monitor the condition of equipment and determine the time of its repair, as well as optimize transport routes and equipment supply schemes.

For example, a data science solution was implementedcompanies Gazprom Neft and Surgutneftegaz. In the first case, the system analyzes equipment breakdowns and predicts when they may occur, in the second, it optimizes business processes and reduces the time required to prepare reports.

  • Agriculture. By 2050, global demand forfood will almost double, so farmers are forced to increase production. Data science will help agricultural producers predict crop failure or, on the contrary, an oversupply of products in warehouses, it is better to calculate the amount of seeds and fertilizers needed, taking into account the characteristics of the climate and soil. This will increase productivity and reduce costs.

In Russia, a similar solution is being implemented by the companyEkoNiva, the largest regional partner of the John Deer agricultural machinery manufacturer. This is a computer vision system developed by Cognitive Technology that turns conventional harvesters into unmanned ones.

Which industries have potential

  • Utility services. In the field of housing and communal services, the analysis of accumulated data will make it possible to predict equipment breakdowns, automatically keep track of the consumption of water, gas and electricity, and also technologize the legal function.

There are only a few cases of implementation in this area. For example, Mosenergosbyt has implemented a data science solution to automate the calculation of charges for overdue payments for electricity. This made it possible to abandon the expansion of the staff of lawyers and reduce costs.

  • Food industry. Big data analysis in the food industrywill help you maintain accuracy in supply planning based on information about the demand for goods, monitor the quality and freshness of products and sell products more profitably through market analysis.

So far, such solutions are being implemented only by large international players - Nestle, AB InBev Efes and others. In Russia, there are no cases of using technology in this industry.

  • The medicine. Big data analytics in medicine will help doctorsmake more informed decisions when making a diagnosis and prescribe the most appropriate treatment; create an electronic patient record from scattered health data and quickly process data from wearable devices without the participation of a doctor.

There are very few examples of the implementation of data science-based products in medicine in Russia - they are used only by telemedicine service providers such as DocDoc and Doc +.

Big data is already the future. Analyzing them will help businesses to better understand their audience and personalize offers for individual customers; collect information about the benefits of the product and evaluate the real user experience; accurately calculate risks and fight fraud, as well as optimize business processes.

If there are industries today that have not penetratedautomation and intellectualization of processes based on big data analysis, then in five years the introduction of data science solutions will become the norm even for the most conservative sectors of the economy.

Big data empowers organizations to reducecosts without a global change in the scheme of work. This gives companies that have implemented such solutions an additional competitive advantage - this is the best incentive for market development.

See also:

Astronomers have discovered a lunar accelerator from the 60s. He's heading towards Earth

Scientist has created a paint that can replace expensive solar panels

The CHEOPS mission has found a planet with the harshest conditions in the universe. the main thing