RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2
2018/12/28 12:13:03

Overview: Artificial intelligence of 2018

Russian market of artificial intelligence: problems and perspectives

In Russia the artificial intelligence (AI, artificial intelligence, AI) already touched upon such subjects as intellectual monitoring of infrastructure, collecting and processing of large volumes of information, knowledge management, technical and medical diagnostic systems, creation of individual trajectories of training, the behavioural analysis, smart platforms, etc. Nevertheless the most serious changes under the influence of solutions on the basis of AI will happen in the next two-three of years.

So, based on the research "Current Trends in the Market of Artificial Intelligence and Machine Learning" conducted by analytical center TAdviser and Jet Infosystems company in October, 2017, the size of the market of artificial intelligence and machine learning (Machine Learning, ML) in Russia it was estimated by about 700 million rubles in 2017. At the same time growth up to 28 billion rubles by 2020 was predicted. New polls of experts confirm high dynamics of the market.

File:Aquote1.png
If in 2017 the number of the Russian projects using AI was calculated by several tens, then in their 2018 already hundreds. By experience of ABBYY, banks and financial institutions are most active in this sphere traditionally, this year "the race of artificial intelligence" joined both the energy companies, and retail, and telecommunications. In the large companies – for example, Sberbank, Gazprom Neft, Severstal, MTS, there were teams of specialists who are engaged only in projects using intellectual technologies. I will assume that in 2019 AI will become already integral part of business of any large company, - Dmitry Shushkin, the CEO "tells Abbyy Russia".
File:Aquote2.png

File:Aquote1.png
According to different forecasts, it is expected that by 2020 the amount of the market will reach $460 million. This trend is confirmed by the investors who selected AI technologies one of the priority directions for investments, - Alexey Tsessarsky, the CEO of IVA Cognitive (Haytek Group) adds.
File:Aquote2.png

Evgeny Kolesnikov, the head "Big Data and Machine Learning", "Jet Infosystems", notices that the forecast read in a research, joint with TAdviser, is confirmed by the whole series economically successful AI/ML projects in 2018 from which about 500 became public.

Igor Kirichenko, the chief executive of Naumen company, notes that his company sees enormous potential in the market of AI and considers that in the next several years the Russian market will develop the advancing rates of rather global market.

It, according to him, is promoted by the following factors:

  • Rather low base of start and overdue growth of this market in Russia
  • Existence of powerful state programs of support of this sphere
  • Presence of local large IT players whose main examination is creation of software products.

Dmitry Chuvikov, PhD in Technological Sciences, the head of department of perspective solutions in the field of AI of Mivar company, lists five main premises of implementation of artificial intelligence technologies and machine learning for the business directions in Russia:

  • The growing need for online services on issue of correct information for the staff of different business divisions (support of the internal and external client).
  • Complexity of high-quality ensuring services of support of business against the background of requirements for decrease in OpEx.
  • Need for management of the accumulated experience and knowledge and also preserving of competences at personnel rotation of any activity.
  • The growing requirements to efficiency of services.
  • Steady trend on intellectualization: didzhitalization of business processes and robotization of work.

In general the companies only begin to understand how to do such projects what are bottlenecks in implementation, their time is much bigger, than in classical projects.

File:Aquote1.png
Customers see the opportunities in the field of implementation of artificial intelligence technologies and machine learning and begin to set tasks to the IT companies. Growth of interest of customers says that the market will increase in the near future, and is quite considerable. Already now we see expansion of number of customer companies which actively use bots or the system of speech recognition, - Alexey Vyskrebentsev, the head of the center of examination of solutions of Foresight company tells.
File:Aquote2.png

As for the domestic IT companies which offer the services in the field of implementation of artificial intelligence technologies and machine learning, their expert experience extends, and opportunities grow.

At the same time many IT companies define for themselves the AI and ML directions as focal for business development in the future. Certain practicians on machine learning open and around it positioning of business not only in the Russian market, but also on international is built.

Temporary shortage of specialists

Transition from universal discussion of machine learning technologies and artificial intelligence to a stage of practical use led to growth of number of Data Science of experts in Russia. However so far in the market still there is a huge shortcoming of similar specialists.

According to Marina Mayorova, the head of artificial intelligence and machine learning Croc, demand for highly qualified specialists in the field of Data Science will continue to grow, however so far the companies experience difficulties searching, attraction and deduction of such employees.

File:Aquote1.png
In 2017 SAP conducted a research of the scientific organizations and projects in the field of artificial intelligence. In the next 5 years not less than 50 thousand specialists will be produced that will surely urge on the market. In our country apparent interest to these technologies both in respect of training, and in scientific developments of the different organizations is visible. In my opinion, every year the size of the market of artificial intelligence and machine learning will only grow because these technologies will gradually cover all industries, releasing time for acceptance of management decisions and creative processes for the person, - Yury Bondar, the deputy CEO of SAP CIS considers.
File:Aquote2.png

File:Aquote1.png
Considering the high level of profile education in our country, the Russian specialists of AI and ML are very much highly appreciated not only houses, but also abroad. It gives the grounds to believe that the domestic companies are able to afford to look at international market, as on potential in terms of strategic development, - Andrey Baybutov, the director of business development of department of BI of Corus Consulting Group is sure.
File:Aquote2.png

Data should make profit

Alexey Vyskrebentsev, the head of the center of examination of solutions of Foresight company, considers that market development of artificial intelligence and machine learning is stimulated with growth of production capacities, increase in the amounts of data which are saved up by domestic enterprises and the organizations and also information on successful cases and on the benefits received by competitors thanks to AI technologies.

At the same time not all companies are able to process data that they worked for advantage to business.

File:Aquote1.png
It is necessary not to accumulate data "as is", and to collect under certain target tasks of business. Their volume, a format, a cut are defined by the algorithm solving a target problem of business. Requirements to data and the connected infrastructure (business applications for data collection, software for storage, etc.) are defined by an algorithm. Such approach allows to find balance between the business value, necessary data and their cost, - Dmitry Karbasov, the head of practice says Artificial intelligence of digital laboratory Softline.
File:Aquote2.png

In CleverData company (Lanit group) note positive examples of application of neural networks in work with user data and solving of tasks of predictive analytics.

Using artificial intelligence technologies it is possible to reveal the hidden patterns in a consumer behavior, to determine the probability of a response to this or that advertizing offer, to understand to whom, when as well as that it is better to offer to build most personalized communications and to increase efficiency of advertizing campaigns.

File:Aquote1.png
The tendency to trust the machine such solutions is proved, first of all by the fact that it is extremely difficult to process a big flow of the generated data on users a human brain. The artificial intelligence copes with a problem of extraction of new knowledge of consumers and the hidden patterns quicker and more effectively, - Denis Afanasyev, the CEO of CleverData notes.
File:Aquote2.png

The competence of customers grows

Many participants of the domestic market of AI fix growth of competence of customers of the questions connected using similar technologies. In particular, Alexander Khanin, the CEO of VisionLabs, referring to experience of the company developing algorithms of computer vision notices that clients and partners began to understand how these technologies work what restrictions have and what they can be applied to.

File:Aquote1.png
It led to the fact that most of clients and partners address for the solution of tasks, feasible for such technologies. In addition, the competence of representatives of the market of methodology and to approach to correct job evaluation of technologies grew, - he adds.
File:Aquote2.png

Dmitry Lukovkin, AI Business Director of Tsifra company, speaks about growth of level of understanding of artificial intelligence technologies and machine learning at production companies.

File:Aquote1.png
In them CDO and departments of digital transformation, along with own departments of Data Science appear. Approach to technologies becomes more pragmatic, - he explains.
File:Aquote2.png

Along with it requirements to suppliers of AI solutions and projects grow. Existence of examination in data domain, both immersion in production, and communication with a first line of science become necessary success factors of the companies working in this sphere.

With trust to AI technologies not everything is smooth

Speaking about trust to artificial intelligence technologies respondents of TAdviser experts were separated in the opinions. For example, in Naumen and IVA Cognitive companies (Haytek Group) notice growth of trust to AI.

File:Aquote1.png
The number of the companies ready to implement and apply at itself AI technologies will promptly grow. Successful project experiment, deeper understanding of opportunities of technology, evidential base by efficiency of application and open results of performance improvement will actively promote it, - Igor Kirichenko, the chief executive of Naumen company notes.
File:Aquote2.png

Alexey Tsessarsky, the CEO of IVA Cognitive (Haytek Group), calls a factor of trust of one of the main trends, and considers that it promotes universal implementation of artificial intelligence technologies.

File:Aquote1.png
Not very long ago, use of AI was limited to research laboratories, now we see growth of interest in this technology, among the companies from the different industries. In the Russian realities of AI it is most demanded in the field of security (the system of face recognition, the ACS), financial (scoring), marketing (the recommendatory systems, chat-bots, outflow prediction), medicine and others. As a positive thing for development of AI technologies and machine learning serves increase in availability of computing powers that does implementation of similar solutions commercially reasonable, - he says.
File:Aquote2.png

The CEO of CleverData Denis Afanasyev, speaks about insufficient trust to AI technologies. According to him, concerns and level consumer confidence to the applied technologies and also single questions of regulation of the market, are market barriers.

File:Aquote1.png
Work with data of Internet users and application them in practice requires confidentiality and safety of personal data. This question is in the center of attention not only the legislation, but also business for which it is important to provide protection of the digital asset, he explains.
File:Aquote2.png

Dmitry Karbasov, the head of practice Artificial intelligence of digital laboratory Softline, notices change of the attitude towards technologies - from construction of artificial intelligence in a cult and a fair share of noise to a certain share of disappointment from powerful neuronets around.

File:Aquote1.png
This disappointment first of all is connected with difficulties of real practical application of their results, - the expert considers.
File:Aquote2.png

Alexey Lapshin, the CEO of Aplana of Business Solution company (Aplana group), also notices surplus of "hype" in the Russian market of AI. According to him, the idea of artificial intelligence still remains opaque to the customer. And effective use of this technology requires a long way to come to two major conclusions – economic and psychological.

The first, economic. AI is an instrument of automatic adoption of linear solutions, such as issuance of credit, choice of the supplier, selection and personel assessment, etc. Therefore no effect of implementation of AI can be reached without replacement or dismissal of the people who are responsible for adoption of such solutions now. But these employees represent some kind of "average" class in the organization - the most extensive, important and rather influential. Be left without these people for top management terrible, unusually and in the nearest future nobody is going to dismiss them.

The second – psychological. In case of implementation of AI to ask and punish for the wrong solution there will be simply nobody simply nobody. Also it is so far very difficult for heads to reconcile to this fact, - Alexey Lapshin explains the position.


Ilya Ashikhmin, the head of scientific laboratory Digital Design, reminds of Gartner outputs concerning development of the various ideas and technologies. They, according to Gartner, go through several similar stages in the development: start, the overestimated expectations, disappointment, education, productive use. This observation is rather universal and is suitable including for the analysis of the occurring trends with machine learning in Russia, the expert of Digital Design is sure.

We can see that many companies tried new technologies of data analysis and found out for themselves that their implementation not such simple process and it is necessary to make rather large number of efforts on its implementation. Here It should be noted that achievement of success in such projects requires rather high level of informatization of internal processes of the company, good analytics in data domain and a sober view on all variety of algorithms of machine learning, including examination of their applicability.

In Digital Design company for the last three years experience of implementation of several projects with use of machine learning and artificial intelligence is accumulated. And we consider that in order that in the market there was no trend on disappointment" in these technologies, it is necessary to estimate soberly at implementations of this sort of projects value and opportunities which use of new technologies can give, - Ilya Ashikhmin says.


Interview with experts

The commercial director of CommuniGate Systems Vladimir Burgov and the leading developer Boris Baranov in an interview of TAdviser told about modern technologies and trends of the market of contact centers.

In structure of AiTeco company in 2018 the Center of cognitive technologies which deals with issues of development and deployment of the systems of artificial intelligence is created. The director of the center, PhD in Technological Sciences Ilya Kalagin tells about industry trends, new channels of involvement of specialists and own development of SmartSel.

Artem Permyakov, R&D-директор Directum companies, in an interview of TAdviser told about the most perspective scopes of artificial intelligence technologies (AI), use of similar opportunities in solutions of Directum and plans of future AI developments.

Who in Russia is engaged in AI and ML technologies

By preparation of the overview TAdviser learned from the domestic companies (and the Russian representatives of the foreign companies) what services in the field of artificial intelligence and machine learning they render. Such data provided 14 market participants. Part of them are highly specialized companies which business is focused on application of similar solutions. A part – giants of domestic IT market for which means and AI technologies are addition to other solutions and services.

Services of the companies in the field of artificial intelligence and machine learning

Company Services
Aplana of group of companies
  • Development of own solutions AI
  • Integration of own and partner solutions AI

Digital Design
  • Research works on studying of opportunities of application of algorithms of machine learning for business process automation.
  • Development and deployment of modules of machine learning in information systems of customers.
  • Custom development from scratch solutions using intellectual algorithms.
  • Integration of platforms of the analysis of databases, implementation of existing solutions.

Jet Infosystems
  • Consulting and development of strategy in AI/ML.
  • Implementation of methodologies of work with the pipeline cost-efficient and paying off AI/ML projects.
  • Custom development of solutions AI/ML.

Croc
  • Monitoring of IT infrastructure based on machine learning
  • Analytics of behavior of users using machine learning
  • Intellectual video analytics in retail
  • Intellectual video analytics for control of visitors and pass of employees
  • Intellectual video analytics for marketing
  • Corporate intellectual assistants (chat-bots + RPA)
  • Predictive and predictive analytics
  • Optimization models for production and logistics
  • Research of data and problem definition
  • Natural language processing
  • Computer vision, recognition of movements and patterns of behavior
  • Deep machine learning

Lanit
  • Line of own software products for automation of communications and online advertizing using technologies of machine and deep learning. Consulting in the field of technologies of processing of Big Data, machine learning, neural networks, predictive analytics and automation of marketing. Project implementation for a telecom, retail, e-commerce, the financial industries and the advertizing industry (AdTech and RTB). (CleverData, is a part of Lanit Group)
  • Analysis and data visualization. (NORBIT, is a part of Lanit Group)
  • Forecasting of sales. (NORBIT, is a part of Lanit Group)
  • Segmentation of the customer base for holding more effective marketing campaigns. (NORBIT, is a part of Lanit Group)
  • Assessment and forecasting of loyalty and vypolneniye of target action by clients. (NORBIT, is a part of Lanit Group)
  • Creation of a digital profile of the client. (NORBIT, is a part of Lanit Group)
  • Custom development in two main directions (The systems of computer vision, is a part of Lanit Group): machine learning in the field of computer vision, machine learning in the field of the analysis of large volumes of data.

Mivar
  • Expert systems of new generation (logically decisive systems, LRS)
  • Semantic technologies of a natural language understanding
  • Intellectual image identification
  • ACS and management systems for independent intelligent robots

Digit
  • Accomplishment of Research and Development and piloting of projects in a scope of machine learning technologies for solving of tasks of the client – from forecasting before generation of recommendations and optimum control in the industry
  • Creation of corporate storages and lakes of data (Data Lake)
  • Consulting on digital transformation of business and monetization of data of production companies

Abbyy
  • Software development. Artificial intelligence technologies of ABBYY allow business and the state organizations to take data from any kinds of the documents (structured and unstructured), to enter information into corporate systems, to analyze data, to perform search in all internal sources of the company, etc.

Directum
  • Development of own solutions on the basis of artificial intelligence
  • Implementation and consulting in terms of automation of routine transactions with the help of AI, digitalization of content and business processes
  • Integration of own intelligent solutions into the customer's IT infrastructure

IVA Cognitive (Haytek Group)
  • Development of systems of video analytics, face recognition, images, sounds.

Naumen
  • Replacement of employees by autonomous AI algorithms
    • - Auto-classification of subjects of addresses with the subsequent intellectual routing (Naumen SMIA, Naumen Intelligent Assistant)
    • - The dialogue robot (a voice and the text) of the centers of client support and dispatching office in OTsO (Naumen Erudite)
    • - A virtual chat consultant on the website and in mobile application (Naumen Erudite)
    • - The robot collector of receivables for B2C (Naumen Erudite)
    • - Specialists robots, for example, on booking of the taxi, questioning of clients, acceptance of indicators of counters of housing and public utilities, providing any consultations and informing (Naumen Erudite)

  • Equipment of employees tools of the complemented intelligence for performance improvement of work and quality of works

    • - Lawmaking and corporate rule-making (Naumen LegalTech)
    • - Management of expert activity with decision support with processing of Big text Data (Naumen Smart Expertise)
    • - Management of personnel potential (profiles of competences, formation of project teams, adaptation and transfer of knowledge). (Naumen Knowledge Cat)
    • - Predictive analytics of failures and repairs of a production equipment and IT infrastructure (Naumen DAP)
    • - Predictive analytics of quality of output products and optimization of a consumption of raw materials (Naumen DAP)
    • - Analytical tools for identification of the repeated events, mass problems, filling of knowledge bases (Naumen SMIA)
    • - Joint work (text editor, comparison of versions, approval, communications). (Naumen LegalTech)
    • - Predictive planning of schedules of linear personnel (Naumen WFM)

  • Creation of semantic search environments for Big text and digital data

    • - Use of the accumulated corporate knowledge (semantic search, personal recommendations, cross-yazychnost). (Naumen Knowledge Cat)
    • - Extraction of "sense" from documents (the informative analysis, establishment of communications, entity recognition, data warehouses). (Naumen LegalTech)
    • - The web portal of the personalized self-service with smart search string (Naumen SMIA)

SAP CIS
  • Delivery of the software.
  • Researches on efficiency of use of machine learning.
  • Implementation services of projects with use of machine learning in business processes of the companies

Smart Engines
  • Recognition of text information in a video flow and on images
  • Mobile and server solutions in the field of image processing and recognition of documents
  • Preparation of electronic images of documents for compact storage
  • Classification of the structured and unstructured documents
  • Machine learning
  • Algorithmic optimization

VisionLabs
  • Development and deployment of a system of face recognition

Own solutions of the companies in the field of artificial intelligence and machine learning

Company Own solutions in the field of AI and MO
Aplana of group of companies
  • Preferentum is the platform of the analysis of unstructured information using artificial intelligence technologies and machine learning
  • Robin is the platform for robotization of business processes and creation of intellectual chat-bots (RPA)

Digital Design
  • Processing of the geophysical data obtained as a result of logging
  • System of optimal placement of vulgar and linear objects
  • The analysis of texts and intellectual document handling in an electronic document management system of SDU "Priority": autocompletion of attributes of documents on the basis of the analysis of the text, autoformation of a route of approval of documents, autoselection of the contractor, detection of links to other documents in the text, intellectual search, etc.
  • Search of anomalies in agreements
  • Intellectual analysis of the legal file of the partner

Jet Infosystems
  • Jet Galatea is the platform for work with data streams and models of machine learning for the analysis of events in real time
  • Jet Detective is the universal solution on counteraction to fraud based on machine learning technologies and artificial intelligence
  • Jet Capacity is the solution for forecasting of requirements to IT capacities depending on business indicators
  • Jet Smart Building – Industrial IoT-solution with AI of a component for management of the engineering systems of retail stores and sales points
  • Jet Signal is the solution with AI of a component for information security incident management: simplifies work of cybersecurity service, arranges processes and saves time
  • Jet Pluton is the system of detection of network attacks and identification of threats on the basis of signature methods and AI/ML technologies

Croc
  • Monitoring of IT infrastructure based on machine learning (User and Entity Behavior Analytics, UEBA)
  • Analytics of behavior of users using machine learning

Lanit
  • 1DMP Data Marketing Platform is the solution for automation of communications and online advertizing using machine learning technologies and Big Data. (CleverData, is a part of Lanit Group)
  • Integration of ML services in the CRM system of clients. (NORBIT, is a part of Lanit Group)

Mivar
  • Wi! Mi (KESMI) – the Designer of Expert Systems Mivarny
  • Tel! Mi (TELMI) is the Text Emulator of the Personality Mivarny
  • ROBO! RAZUM (ROBO! RAZUM) – a software platform of intellectualization for robotics

Digit
  • Zyfra Eye – A product for recognition of dashboards of a production equipment using CV and Deep Learning
  • Zyfra CV – The solution for recognition and calculation of industrial facilities and personnel using CV and Deep Learning
  • Zyfra vSensor – Solution for indirect determination of qualitative indexes of products of chemical production
  • AI Heat Treatment – Solution for process optimization of heat treatment of pipes
  • AI Smelting – the Solution for end-to-end process optimization of smelting became in electrosteel-smelting workshop
  • Robotic loading transport system "Intellectual pit"

Abbyy
  • ABBYY FlexiCapture is the universal platform for intellectual information processing with machine learning technologies.
  • ABBYY Intelligent Search is an intelligent solution for quick search of data and documents in any corporate sources.
  • ABBYY InfoExtractor is the solution which with a high accuracy takes information, important for business, from data arrays.
  • ABBYY of Smart Classifier is the tool for document classification which allows the organizations to distribute automatically a flow of incoming documents, to instantly find or archive information in corporate systems.

Directum
  • DIRECTUM Ario is a set of the intelligent tools using algorithms of machine learning for disposal of routine transactions in an ECM system:
    • - processings of a flow of documents and their classification,
    • - extraction of significant data,
    • - automatic registration and routing,
    • - information search, etc.

IVA Cognitive (Haytek Group)
  • Video analytics
  • Face recognition
  • Image identification
  • Recognition of sounds

Naumen
  • Naumen Knowledge Cat is a knowledge management system, competences and information flows. Clients: large and medium-sized companies (in particular, HR services).
  • Naumen LegalTech is the system of intellectual processing of legally significant documents. Clients: state. bodies, large and medium-sized companies (legal services, rule-making, corporate management).
  • Naumen Service Management Intelligent Automation (SMIA) is tool kit for control automation of IT and customer services on the basis of technologies of Big Data and machine learning, allows is automated to solve problems and to make recommendations to specialists of customer services and end users, reducing the cost of support of systems and eliminating inefficiency of processes at manual data processing
  • Naumen Semantic Responder is universal dialogue interfaces for corporate information systems
  • Naumen Erudite is the AI platform for creation of voice and text robots for remote mass service
  • Naumen Contact Center is a communication framework for automation of mass remote service
  • Naumen Workforce Management – provides continuity and succession of process of management of labor resources from forecasting of loading before reporting on use of working time.
  • Competence center on intellectual automation
  • The Center of researches, joint with NITU "MISIS", in the field of Data Science

SAP CIS
  • Tools for business of analysts: SAP Analytics Cloud with functions of the smart forecast and automatic search of insights; SAP Predictive Analytics for fast creation of forecasting models, classification, the analysis of interrelations, etc.
  • The Data Ops tools for implementation of intellectualization of business processes: product line of SAP Leonardo, SAP HANA, SAP Data Hub.
  • The AI and ML elements which are built in a digital core of SAP: SAP S/4HANA, SAP C/4HANA, SAP SuccessFactors, SAP Ariba.

Smart Engines
  • Smart IDReader is the system of artificial intelligence for recognition of documents

VisionLabs
  • The Luna platform for the analysis of persons of people, photo and video these with the image, for the subsequent comparison with databases


Solutions of third-party developers in the field of artificial intelligence and machine learning with which the company works

Company Solutions
Aplana of group of companies
  • UiPath is the RPA platform

Digital Design
  • Open source of the solution based on which the scientific laboratory "Digital Design" is engaged in writing of own software products focused on accomplishment of specific objectives.
  • Develops and implements solutions based on Microsoft Azure and Power BI.

Croc
  • Microfocus UBA
  • Vision Labs
  • Rapidminer

Lanit
  • OpenSource of the solution, for example: Catboost (Yandex), TensorFlow (Google), Scikit-learn packet. (NORBIT, is a part of Lanit Group)
  • Azure ML Studio (Microsoft). (NORBIT, is a part of Lanit Group)
  • The company makes custom solutions under needs of the specific customer, at the same time in 100% of cases open libraries and solutions are used. (The systems of computer vision, is a part of Lanit Group)

Naumen
  • The systems of synthesis and speech recognition (ASR\TTS) from such vendors as Yandex and CST

SAP CIS
  • SAP uses possibilities of virtualization from Google Kubernetes for containerization of any open source of algorithms of ML.
  • Deep integration with TensorFlow framework.
  • Use of the most popular languages for data analysis: Python, R, SQL.
  • Integration with Hadoop.

The Russian projects in which are applied artificial intelligence technology and machine learning

Company Projects
Aplana of group of companies
  • The Ministry of Internal Affairs of the Russian Federation is development and deployment of a system of legal monitoring (AIS "Monitoring") and the system of legal examination (AIS "Monitoring-M") based on the Preferentum platform
  • Aktion-MTsFER - creation of a cloud online service of intellektalny processing and the analysis of agreements based on the Preferentum platform
  • Freight One is development of the program robot "Analysis of Delays of the Rolling Stock" based on the Robin platform
  • Russian Post - the program Avtosverka robot for the analysis of financial data based on the Robin platform

Digital Design

Jet Infosystems
  • The system of prediction of behavior of buyers for Rive Gauche
  • Transport control system for Kontrol Leasing
  • The system of commodity recommendations for network of drugstores "the Century Live"

Croc
  • Smart video surveillance in the Sakhalin region
  • The managed Croc service based on technology of computer vision (Metro Cash and Carry)
  • Chat-bot "Croc"

Lanit
  • Optimization of communications with clients of online stores of goods using machine learning technologies. (CleverData, is a part of Lanit Group)
  • A recommendatory system for service of armoring of tickets using neural networks for determination of consumer interests and the direction. (CleverData, is a part of Lanit Group)
  • Development of analytical models on the basis of data on consumers for determination of probability of a response and outflow. (CleverData, is a part of Lanit Group)
  • Algorithm elaboration of ADAS (System of the help to the driver)
  • The Amur tiger – the solution on checking and identification of tigers on the drawing of a skin
  • Selection of heart and its valves according to pictures of ultrasonography

Mivar
  • Implementation of the intellectual chat-bot in Alfa-Bank

Digit
  • Optimization of a power consumption at network functioning of chillers "Filip Morris Izhora"
  • Development of an analytical system using machine learning for workshop of production of chlormethanes
  • Gazprom Neft and Tsifra company signed the agreement on cooperation when implementing the Digital Plant project

Abbyy
  • The VTB bank automated opening of the account for legal entities using artificial intelligence technologies of ABBYY.
  • The Tochka bank automated processing of addresses to client support using AI.
  • MUEC integrated accounting and financial control on the universal platform with the AI elements.
  • NPO Energomash implements search based on AI of ABBYY which will integrate millions of documents from 9 corporate data sources in the general system

Directum

Naumen
  • Voice robots on acceptance of meter readings for Interrao
  • Voice - the robot the qualifier and robotic service "Where My Sending?" for Russian Post
  • The project in the field of machine learning for STC Gazpromneft

SAP CIS
  • Loyalty program "my Victoria"
  • Project on forecasting of a burn-through of tuyeres for NLMK
  • Analysis of the bank statement for NLMK

The largest projects of SAP in the field of AI not of a publichna.

Smart Engines

VisionLabs
  • Development of a Single biometric system
  • Development and deployment of a system of face recognition and access control in Sberbank branches.
  • Development and deployment of a system of face recognition in network of the Father cafe Jones and Media cafe for payment of the order using a selfie

Trends of the market of AI and ML

Replacement of employees with robots

Any repeated dialog – a workplace for the robot. As a result modern Russian contact centers actively implement robotic services for processing of client requests, thereby improve service quality.

File:Aquote1.png
If rates of acceptance of robots for work to remain, then in 5–10 years operators of contact centers will carry out no more than 20% of the arriving tasks – all the rest will be done by robots. Only in 2018 Naumen robots serviced more than 5 million dialogs, - Igor Kirichenko, the chief executive of Naumen notices.
File:Aquote2.png

AI combination to robotization of business processes

One more trend – AI combination to robotization of business processes (RPA). In this linking of RPA performs the simple repeating operations: transfer data from one system to another, open and close programs, send automatic answers.

File:Aquote1.png
AI gives to robots additional skills and helps to recognize texts, to define document type, to take significant data that then to transfer them to direct information systems. Similar combinations of technologies already implement large banks, the energy companies, retail networks and other participants of the Russian market, - Dmitry Shushkin, the CEO "tells Abbyy Russia".
File:Aquote2.png

Expansion of opportunities of software products means of AI and ML

Speaking about technology trends of the Russian market of artificial intelligence and machine learning, it is worth mentioning increase in computing opportunities and also expansion of functionality of software products in a support area of the most different data sources and recognition of a natural language.

File:Aquote1.png
For example, the Foresight company actively develops functionality of the products in the field of machine learning: we integrated internal language of our developments with the Python language which "is ground" for solving of tasks of Deep Machine Learning. Thus, effective algorithms of machine learning for the most different purposes facing the enterprises from the different industries allow to create our solutions (in particular, it is about product capabilities "Foresight. Analytical platform"), - Alexey Vyskrebentsev, the head of the center of examination of solutions of Foresight company tells.
File:Aquote2.png

New methods of machine learning

One more of the main technology trends – new methods of machine learning which accelerate development and projects implementation in the field of AI in the conditions of the small number of data. It is, for example, transfer learning. The method is that if to train deep neural network to carry out one task, it is possible to use the same architecture of a neuronet for training at other data set.

File:Aquote1.png
We use such developments in the solution ABBYY FlexiCapture: previously we train network for document classification at the party that on the party of the customer the algorithm could begin to work with the minimum quantity of the marked documents, - Dmitry Shushkin, the CEO "tells Abbyy Russia".
File:Aquote2.png

Wide use of AI for service quality improvement

The following trend is connected with wider use of AI for service quality improvement, both online, and offline. In digital channels it is, first of all, collecting and information analysis about the buyer or the user through "a digital trace" - a data set about a customer behavior on the website, in mobile application, in social networks, it is on purpose better to understand it and to offer it the most suitable goods and services.

Offline it can be sensing technologies of persons, identifications of objects and the behavioural analysis. In particular, systems which are preventively guessing intentions of people are already created.

In addition, similar solutions help to make service in sales points more address: define gender and age of the client, and in a case with frequenters – detailed information on the buyer that allows to understand more precisely what products will be interesting to it at present.

File:Aquote1.png
Besides, such technologies give the chance more simply and quicker to provide services, including in the self-service mode – for example, to make out the SIM card or to buy tickets for transport, - Dmitry Shushkin explains.
File:Aquote2.png

Universal use of AI

The global trend is that all developments and solutions in the field of AI anyway will be will be applied in various areas. But the result will differ from such application on effect scale.

File:Aquote1.png
The artificial intelligence already is at the level of universal software solutions and mechanisms which it is possible to turn on in the services. It is definitely possible to tell that the artificial intelligence will cover all IT services, - Igor Kirichenko from Naumen tells.
File:Aquote2.png

Where use of AI is most relevant?

The artificial intelligence and machine learning appeared in the middle of the last century. In recent years, thanks to emergence of new fundamental techniques, availability of huge volumes of the saved-up data and development of hardware capacities, interest not just returned, and got on the peak of technologies in the development scraper from Gartner. It means that now time of the highest expectations when the artificial intelligence is often considered "a silver bullet", try to use in all areas, even in where it or is not necessary at all, or it is possible to manage by traditional algorithmic methods.

File:Aquote1.png
Weekly we see emergence of new impressive projects based on machine learning in different areas: to speech processing and images, auto pilots, etc. Probability is high that in 2-3 years information noise around Machine Learning will fall down, and there will be a set of examples rational use of these technologies allowing the companies to get strong competitive advantage, - Dmitry Timakov, the head "Machine learning" of NORBIT company (Lanit group) believes.
File:Aquote2.png

According to him, Russia has good chances of inclusion in the world market of AI. In labor market there are excellent specialists (at a top of the main competitive platform kaggle steadily there are Russians), there are examples of creation of world-class products: for example, the voice assistant Alice from Yandex company.

File:Aquote1.png
Today we see that large business (IT, retail, etc.) is opened by the internal divisions for implementation of machine learning technologies. There are several spheres where the artificial intelligence can be integrated into business processes most successfully, - the expert of NORBIT notices.
File:Aquote2.png

Such spheres, according to him, treat:

  • personalisation of offers to clients, creation of relevant recommendatory services of new generation;
  • hands-off processing of the user content and actions: for example, analysis of responses, addresses, identification of bots, etc.;
  • creation of elements of concepts "Smart and safe city": from prediction of traffic jams before intellectual management of traffic;
  • the consulting help to doctors when carrying out medical researches;
  • industrial automation in the industries, electronic vision systems, etc.

Dmitry Chuvikov from Mivar company the main and first-priority consumers of intelligent systems calls the business directions with the following indicators:

  • Set of different transactions in processes of rendering services.
  • The structured and formalized business processes.
  • Complexity and a variety when choosing options of actions.
  • High level of standardization of processes.
  • Need for information and consulting support here and now.

Thus, according to him, the scopes of artificial intelligence technologies and machine learning, most relevant for Russia, are: banking sector, retail, insurance companies, telecom, oil industry, military industrial complex, pilotless transport and "Smart city".

In the financial sector of implementation of the systems of class AI the fintekha stops being a prerogative only. It is more and more classical banks and insurance companies understand an economic case of efficiency of use of AI in their processes and create financial products using machine learning and computer vision.

According to Vladimir Arlazarov, PhD in Technological Sciences, the CEO of Smart Engines, in the financial sector of AI technology are applied to creation of new user experience and the analysis of Big Data. This results from the fact that at banks and insurance companies many data are saved up and their business is directly connected with identification of dependences between them. Already now neural networks take part in decision making about approval of the credits and calculation of cost of an insurance.

File:Aquote1.png
As for computer vision, from the market there is already a request for the solutions providing industrial quality of recognition of labels, markings on parts, calculations of goods, etc, - Arlazarov notes.
File:Aquote2.png

He adds that when queue reaches industrial enterprises there will be the next serious break.

File:Aquote1.png
They can save huge amounts of money, for example, predicting using AI of breakdown of the equipment. In Russia this task so far to solve rather difficult because the industry at first needs to collect data. In the world successful projects based on computer vision and machine learning in the industry already are, - the expert of Smart Engines notes.
File:Aquote2.png

Andrey Baybutov from Corus Consulting, adds that use of machine learning technologies in the industrial companies involves strong change of internal business processes without which achievement of notable result is impossible.

File:Aquote1.png
I will give a simple example. The industrial company has a number of difficult technology processes which can be broken into to small stages / subprocesses. At each such stage we analyze controlled parameters at "input", we check a product which is given to the next stage regarding quality. Even if process of production quite simple, engineers cannot physically control in each unit of time what happens to change of parameters (temperature, pressure, % of the contents and so forth). Using AI we can create models on each of stages, each of which controls all set of parameters throughout all process of production of goods and shows bottlenecks. It allows to expect probable emergence of defects in a batch. Having such information, the company can eliminate all defects still before it occurs that of course, taking into account the cost of the potential missed profit allows to save essential money, - Andrey Baybutov tells.
File:Aquote2.png

For this reason the market of the industry begins to move slowly towards digitalization of processes not only in terms of implementation of systems, but also use of change of processes and level of competences inside. In the companies there are analysts of data and developers of models, data-offices are created. Business begins to understand that income which is gained by the company can form not only thanks to primary activity, but also thanks to work with data.

Igor Kirichenko, the chief executive of Naumen, referring to experience of the company notes that in Russia the dialogue systems and services are demanded: voice robots, chat-bots.

File:Aquote1.png
The robotic services capable to conduct dialogue with the client are successfully implemented and are used by the companies of the different industries. In a portfolio of the company there are projects including the innovative developments on the basis of technologies of natural languag processing, machine learning and mathematical linguistics: self-service portals, solutions on cognitive search which process thousands of documents from different sources and allow to find easily the necessary information, there are also successfully implemented projects in the field of predictive analytics - solutions on intellectual monitoring of IT infrastructure, - the representative of Naumen tells.
File:Aquote2.png

The success of use of artificial intelligence technologies depends mainly on presence of sufficient amount of data which are required for effective training of analytical models. Especially it affects results of work of models of deep training – neural networks, Denis Afanasyev, the CEO of CleverData considers.

File:Aquote1.png
Such spheres of business as IT, telecommunications, finance and also online media can brag of the large volume of user data which deep analysis allows to solve problems of client analytics and personalized marketing, - Afanasyev says.
File:Aquote2.png

Data on activity of each user on the website, in mobile application, data about online and offline sales, reactions of the customer base to communications in different channels – from all these data the machine is capable to take new useful knowledge of interests, intentions of each consumer. It can also predict the probability of a response to communication and the advertizing offer, understand which of clients is most loyal who needs to be pushed a little to purchase and who can refuse offers of a brand in the nearest future and go to outflow. Further all this knowledge allows to optimize communications, online advertizing and to offer clients the new level of service.

Around the world B2C-industries the most sensitive for use of artificial intelligence and machine learning. And Russia here not an exception – the telecom, retail and finance make the main industries setting speed in a scope of AI and ML today, Yury Bondar, the deputy CEO of SAP CIS notes.

Besides, It should be noted separately the metallurgical and mining industries which also because of the high competition in the market are forced to look for constantly new approaches to transformation of business processes and to use the innovative technologies.

File:Aquote1.png
We in SAP keep the statistics of use of AI and ML technologies in different business processes of the companies: about 25% are the share of production, 28% - for sales, 13% for HR, 26% for finance, 4% - of purchases and 4% – of all the rest, - Yury Bondar explains.
File:Aquote2.png

Actually any industry where the large volume of data is saved historically up, already uses new technologies. For example, for search of the candidate for the company the machine can quickly make primary selection from the sent summaries. As input data it needs to set only wishes to future employee. It will find and will process a profile of the ideal candidate, further will compare it with data on social networks. At the exit the HR manager will already have a ready profile of the candidate, and he will need only to make the decision on need an interview. Or processing of the appeal to the company: the artificial intelligence can classify a request for a long time and to redirect it to the necessary support service and also to answer standard questions.

File:Aquote1.png
One more option is image identification. The BMW company often wages advertizing campaigns in the different countries and using SAP Brand Impact estimates the market regarding emergence of a brand on different platforms in advance: on billboards, in movies, in videos. As a result the detailed report forms, what is the time in air and in the center of attention of the person the machine in a profile, a fullface, sideways, on top appears. Of course, the person cannot make it. And at stadiums during the football matches the system of video recognition on the basis of artificial intelligence with an accuracy of millisecond will estimate where it is necessary to show advertizing on what period and will express it in a cash equivalent, - the expert of SAP CIS tells.
File:Aquote2.png

Good example of use of artificial intelligence is automatic acoount reconciliation and the arrived payment documents in the financial block of the organizations. For example, in the Alpiq energy company thus more than 90% of the arrived payment documents are verified. Besides, it is possible to apply artificial intelligence and in the national economy to visual control of important nodes of the equipment which is in difficult and dangerous zones of production.

Alexander Khanin, the CEO of VisionLabs, notes that the requests for the algorithms automating process of control of observance of the security regulation, simplifying criminalistics, personnel records and the other at all tasks which are not connected with each other began to appear. At the same time they perfectly are solved using artificial intelligence and machine learning. In a priority, according to him, there are such scopes of application as image identification, persons, a skeleton, patterns of behavior, a voice and the text.

The high demand in the systems of video analytics and face recognition, is noticed also in IVA Cognitive company (Haytek Group). According to the CEO of the company Alexey Tsessarsky, these systems based on work with neural networks gradually become an integral part of our life.

File:Aquote1.png
It is explained by existence of huge data array - daily, at least several times a day, we anyway get to lenses of video cameras, the value of these data does not raise doubts both for business, and for the state, - he says.
File:Aquote2.png

Evgeny Kolesnikov from Jet Infosystems notices that for the last year in Russia such scopes of AI as work with clients (individual offers, segmentation, scoring, optimization of a product matrix, sensitivity assessment to the price, etc.), work with personnel (reduction of idle times, optimization of the diagram, control of contractors, etc.), optimization of production processes (prevention of defects and accidents, optimization of warehouse stocks and supply chains), ensuring information and physical security were brightly traced (identification and prevention of theft of clients, suppliers, employees, protection against penetration on physical and information objects, etc.).

According to Ilya Ashikhmin, the head of scientific laboratory Digital Design, artificial intelligence technology and machine learning find broad application in automation of production processes, the analysis of legally significant document flow, solving of tasks of video analytics, increase in marketing effectiveness and advertizing and in other numerous areas – beginning from scientific research, finishing with the entertaining sector of economy.

File:Aquote1.png
If to speak about specifics of Russia then It should be noted that owing to the big sizes of our country and urgent need in development of infrastructure, the most perspective, in our opinion, scopes of artificial intelligence are problems of logistics and capital construction which allow to reduce considerably costs for the solution of these tasks. In addition, there is a big request for increase in efficiency of the enterprises due to digitalization of decision making processes, being based on the analysis of a huge number of information which is possible only using modern optimization methods, - he adds.
File:Aquote2.png