Alexey Zaslavsky, AiTeco Group: On the platform of "The digital alphabet" – to the digital world

Alexey Zaslavsky: Data are blood of digital economy, heart of any big and small digital transformation. Exaggerating a little, but not changing the essence, one may say, that any fashionable digital technology is a technology of generation, transfer, storage, processing and extraction of data and/or receiving using these data of a certain value.

Avoiding declarations on the impressive growth of the information flows generated and processed in the world in recent years, I will note only that this process is connected first of all with growth of volumes of unstructured or semistructured data. Also requires use of the specialized tools and technologies differing from the instruments of work with structured data which were earlier classical, saving at the same time relevance of effective work with them.

Let's imagine a certain pyramid which levels from the bottom to top can be described as data – information – knowledge/value. On each of levels it is necessary to solve a complex of the interconnected problems using different techniques, methods and technologies which should be directed to achievement of an ultimate goal – extraction of required knowledge and/or obtaining necessary value.

The range of such tasks is wide: from determination of data sources, mechanisms of their generation, collecting and updating, ensuring effective storage and data access – through solving of tasks of processing, ensuring the necessary quality of data, the organizations of work with meta - and master data, – before transformation of data into significant information at the expense of their horizontal and vertical kontekstualization, use of the different tools and technologies directed to obtaining required knowledge and/or values.

Alexey Zaslavsky: Including. Big Data in general are provided by many answer to all questions today. We take data, we start a script and we expect that we will instantly receive the idea of a disruptive innovation or understanding at the exit how to change a business model. And when for various reasons it does not occur, it causes bewilderment or gives rise to mistrust.

Not Big Data as such, but "smart" data which are competently picked up and processed from which it is possible to take required value are necessary. Also their volume, how many understanding of a specific purpose which are correctly selected and the used tools which will be able effectively to work with each of three V-components where volume is amount of data is important not so much, velocity is the speed of gain/processing, variety is variety of data.

Alexey Zaslavsky: Spoke about 10-12 years ago about scrappy automation when for the solution of specific objectives the new information system was created, and then long-term and expensive programs of elimination of duplication and transition to a target IT landscape were developed. Today the companies can go (not) purposely on the way of "scrappy digitalization": solve local problems due to point implementation of innovations, without having learned to work with such new asset of digital economy as data effectively. "Digit" of it will not forgive: speed of changes increases, and with it also the price of an error of inefficient use of own assets, i.e. data grows. It is important to understand that others will already correct this error...

Let's consider the organization of so-called "these lakes" – storages where merge all data of separate branch in hope to extend from them then an advantage maximum. Or, on the contrary, to one lake data of all divisions of the company on the principle "all to the house" merge.

In the first case, whatever mighty analytics was attracted to data processing, its outputs will be them and are limited. There is a risk it will be impossible to receive many small ponds and dams where nothing is more essential to catch small crucians even if will apply the latest artificial intelligence technologies to their catch, and for accounting – a blockchain. In the second option, for lack of mechanisms of management of metadata, accurate procedures of collecting and updating of data, the big lake can quickly turn into the swamp where bloodsuckers yes of a frog" are found some ".

It is necessary to create the whole hydraulic system with the system of water reservoirs, gateways, channels, the systems of cleaning using which it is possible to receive the necessary products and services: catch fish, irrigate fields, boat, develop the electric power, deliver to consumers service or drinking water.

Alexey Zaslavsky: As it is banal, it is necessary to provide balance for the fast solution of industry tasks on the basis of methodically competent approach which will not bring in a result to scrappy digitalization. Therefore now to the forefront there are platform solutions capable to provide as flexible work with data of different types, and implementation based on this platform of specific industry solutions with a possibility of their scaling in the long term.

Alexey Zaslavsky: It is a little companies which are given rise digital. Like those can be considered brands of Uber, Airbnb, Netflix. The others are forced to digital economy to adapt. Generally large enterprises have a certain infrastructure in the form of the equipment and other assets, the current production problems, schemes of interaction with clients and partners and traditional architecture of automation today.

The most boldly infrastructure tasks are visible at the enterprises of the real sector. Solutions of a class PCS or SCADA at the bottom level existed long ago. They collected and processed primary data that in a certain measure brings together them with modern ideology Internet of Things. These data, however, generally for control of products and processes on the most critical sections of production were used.

The current technology trends are like that that more and more industries will begin to rise by the level of the digital production and providing infrastructure: trade, logistics, real estate administration. Different tags, the systems of global and local positioning, sensors of physical parameters and other components of smart infrastructure will gradually form a basis for creation of powerful industry infrastructures of collection of information "in fields". Somewhere the layer of primary data collection will be more directed to increase in efficiency of internal business processes, somewhere – processes of relationship with clients and partners.

The low layer of automation will develop in the industries too. Repair of the equipment on a status more often is recognized as the most effective method of service, and its success is directly connected with processing of a large number of primary data.

At the top level of management (and automation, respectively) we observe rather universal picture in an industry section today. In functional managerial "wells" the information flows are directed generally in vertical direction: from the management of division to subordinates and back. In addition to classical business transactions, we can obtain information reflected in documents, master data, historical information (for example, about repairs of the equipment). However it, as a rule, does not form a uniform picture of data.

Alexey Zaslavsky: If to speak extremely simply and briefly: the common information space connecting all available data is necessary. It is base for solving of tasks of the digital economy which is completely based on data. Data really exist separately. Even if information borders between functional wells of the company tried to overcome at the expense of process (read – horizontal) managements, completely a problem earlier it did not solve. At best it could work in the available processes, but in digital economy it is necessary to create new. Besides such, purely process in fact, mechanisms of data exchange are not really well combined with event model.

The fact that the corporate information resource as some kind of synergy of data, different business systems can play a significant role in business of the company and even to become one of its key assets, realized even before digital transformation. The concept of Data Governance on which often advise the large companies resulted to pay attention. Being the "heavy" concept of a uniform view of data, it is more expected total penetration into the culture of business, it is followed by considerable number of rules, organizational conversions, the entered positions and roles. In practice of the enterprise most often solve separate business challenges for what they need a bit different tools. And the concept of Data Governance "matured" during a dotsifrovy era. Anyway, to cope with new calls, other tools and approaches are necessary for business.

There are high-quality shifts: data collection is not tied only to business transaction, production operation, a working shift, the client's visit any more. Data are generated and transferred constantly. Many of these flows "flow" continuously as began the rivers, without having in time, the end. It is a call for the modern enterprises and a question of existence of methodology and the tool too.

Alexey Zaslavsky: During a dotsifrovy era there was a set of integration products: integration buses, class middleware products. Being successfully used and until now, they are focused one integration more classical business systems and aimed at work with structured data, not in the best way fitting into digital economy where everything consists of not the simplest from the algorithmic point of view, but at the same time rather local from a position of the business importance, tasks.

Their importance that they are solved constantly and in a set. At the same time solutions can integrate in certain pipelines, purchasing bigger weight for business. It is almost always a unique combination of the functions connected, say, with object recognition, a clustering, classification, prediction of their behavior. Attempts to find uniform effective model of processing of data for the organization, to implement it, and then to work according to it for years, gradually debugging nuances, – all this definitely not about digital transformation. Here it does not happen in principle.

Event focus – one more significant postulate of digital economy which is directly connected with management on the basis of data which can be any: structured or unstructured, single or stream, in the form of separate numbers or multidimensional arrays, internal or external, the top or bottom level of management. Integration between them should be performed not on layers (for example, from the PCS level – through the different managerial systems – to a business intelligence of the corporate level), and according to any scheme dictated by a specific objective or an event. In this sense the data on the current temperature and humidity, documentary archive with information on breakdowns of the equipment have no as earlier, a pronounced hierarchical structure.

Alexey Zaslavsky: IT support of digital economy is not business applications in usual sense, and platforms. We start a certain pipeline on data processing – from their collecting on different sources through cleaning, conversion and consolidation before processing with the necessary algorithms and if it is necessary, issues in external systems. It is characteristic that it becomes not in the visual environment, and mainly by means of writing of programs. It could and be lowered as a certain technical feature if business in the majority was not negatively configured to programming in the purchased business solution and would not consider that the system provider offered it to the customer, in it having left unfinished something. Let's tell fairly: proceeding from the previous experience, the business customer has the right to such opinion. But for solving of tasks of digital economy on client side not to do without programming.

Market offers are divided into several categories. A number of the functional systems according to the standard and checked scheme can be used already now. For example, to adjust management being in runs by transport on the basis of the data collected in real time. But out of limits of a limited circle of tasks, let and important for the industry, with their help not to leave.

The most versatile tool of solving of tasks of digital economy also the stack of the products Open Source, known in the market, seems quite developed technology. However serious competences and human resources are necessary for full mastering of this powerful and flexible set of technologies.

There are commercial products which supplier can offer additional quality, say, concerning the solution of widespread tasks or a usability. At the same time it is necessary to understand that "under a cowl" such products, as a rule, have same Open Source not always advertized by the company that it is impossible to consider deception of the customer. It is quite legal and civilized practice.

Alexey Zaslavsky: Recently the company Witte Consulting" AiTeco (Group) brought to the market the Digital Alphabet platform capable to work with any data types in the most "democratic" mode, i.e. without tough "bureaucratic" approaches to separation of data into managerial layers, and with an opportunity to quickly create the combined chains of collecting and data processing of any sources under a specific objective. Record of the processed data is also performed in different databases. Let's tell, we can write the document in the document-oriented database Mongo DB, telemetry – in Cassandra, the graph – in Neo4j, classical transaction – in a SQL database.

Once you note the high computing and communication potential of "The digital alphabet" with implementation of any algorithms of data processing characteristic of digital economy, and a possibility of creation based on the platform of industry solutions. If the customer considers it expedient to solve a business challenge in other system (and with the powerful park of IT systems in some cases it can be more preferable to clients), our product should give collected data to this system and if it is necessary, to take away them back for further processing.

At the same time "The digital alphabet" maintains ample opportunities on connection of dedicated components of work with data: stream processors, specialized databases, tools for predictive analytics and machine learning, allowing to receive necessary knowledge/value from data or with their help. The efficiency of the platform is caused also by the fact that all principal components are implemented on the basis of the mathematical model of parallel computings which is under construction around a concept of the actor.

Submitting "The digital alphabet", we do not hide that in addition to own practices as a part of the platform many products of storage and information processing from Open Source stack – civilized and already established practices for the business companies making software lie. If we sold "box", we could be reproached with attempts to profit not on the work. But in a class of those platform solutions to which also ours belongs "The digital alphabet" the concept of a shrink-wrapped software product in principle does not exist. We offer own developments, experience and examination of our specialists and partners implementing projects based on this platform in a complex, giving to the customer the effective tool and helping it with adaptation under specific business challenges.