‘’Today each IT company is to become a micro Microsoft”
The IT industry is the leader in terms of R&D investments. What changes did occur in IT field recently?
We are going through the 4th industrial revolution which is characterized by explosive changes in all stages of information processing. Indeed, information generation was influenced by the advent of social networks and the Internet of things as both generate data amounts that grow exponentially. Every five years, the amount of information generated in the world increases by an order of magnitude. Data transfer technologies are revolutionized. The technology of verified information exchange, i.e., blockchain, emerged, as well as innovative approaches to distributed storage and parallel processing of data. Machine learning and artificial intelligence help isolate new entities fr om information, with virtual and augmented realities drastically changing its visual presentation.
In what way do these changes transform the market operations of customers, vendors, and system integrators?
15-20 years back, innovations used to come into being in R&D labs of the information giants such as IBM, Oracle, Microsoft, and SAP. Most innovations followed the same route, fr om the lab to the end customer: product documentation was drawn up, staffs were trained and afterwards trained partners, the partners brought the technology to customers trying to steer them into understanding its value. Eventually, the technology was just another product in the integrator’s arsenal and was employed by a wide array of customers. The process itself took fr om three to eight years. But at some point it became clear that it took too long. It will be recalled that in 2013 Bitcoin was only known to domain experts, geeks, and criminals. In another year, the professional community became aware of it, and in two years Bitcoin became publicly known. Now, only five years later, the popular appeal of crypto-currency is on a decline for certain reasons.
So, everyone have to respond. Vendors thus change their attitude to the sacred cow of copyright, right in front of your eyes. A decade ago, one third of any contract with a foreign vendor used to stipulate intellectual property protection minutely describing what may or may not be done. Now, all vendors make their tools publicly available as free products with open source code. Microsoft launched Azure with many free products; Google and Facebook published their TensorFlow and PyTorch libraries for operations with big data, while IBM made its HyperLedger platform available. These tools are available to any programmer. Now, it is borne in upon vendors that fencing off their copyright does not make so much sense as placing their goods on the market to enable partners and programmers to use those products, bring them to the masses, and develop new technologies.
The next link in the chain is a system integrator each of whom now ought to have an R&D lab of its own and morph into a mini Google of a micro Microsoft. When working with cutting edge technologies, there are no ready answers, and each project requires first-hand experience in the data of a specific customer.
Crucially, the role of the customer reshapes itself and becomes of utmost importance. Only the customer knows where and how data is to be mined. Next, data are processed by the system integrator and presented to the customer. Hold onto your hat, in 99% cases the first result does not work for the customer, which means you have to iteratively, step by step, retrace your data processing. The only criterion is your customer who determines if the result is positive or what is to be done to improve it.
I would like to introduce an example of the customer role in assessing the results. At one time we set up a system for processing incoming information from different communication channels for an airline. The goal was to automatically detect the tonality of messages and to determine if the tonality is positive, or neutral, or negative. We built a model which was highly accurate according to all classical rules, but the customer was aghast claiming the model was irrelevant. When we asked for an explanation, the customer gave us the following example: “Look, there is a girl who writes how she bought a ticket for a flight, left for the airport, got a flat tyre, was caught in the rain while retiring, was late for her flight, got drenched, caught a cold, and is ill in bed at home. Your system believes it is a negative message”. Our response: “Well, yes, being late for one’s flight and falling ill is negative, after all”. “No, it is not”, says the customer, “It is negative for the girl, but it is not the fault of our airline and we consider the message as neutral”.
What are the innovation development trends? And which novel technologies are most promising, in your opinion?
The popularity of all novel technologies develops classically, as a curve with a hump and tail. When a novelty hits the market, everyone is eager to test it. Start-ups mushroom and trail breaking customers implement the technology. In the course of time, it turns out that the technology is far from being always economically feasible, and the interest is on the wane to some extent.
But at some point a niche is found wh ere the technology is economically effective. The popularity rebounds, but more slowly, until it freezes at certain point.
I can see two booming technologies. The first one deals with big data analysis, machine learning, and artificial intellect. Another technology is Process Mining enabling us to analyze and detect vulnerabilities in a company’s business processes on the basis of a customer’s information systems’ data. IoT and Blockchain have already a good track record and are growing steadily.
In which business lines are innovations most efficient?
They are at their best wh ere there are big and ever-growing volumes of data. If data systems are complex and heterogeneous, if unstructured and incomplete data come from different sources, the analysis does not come easy to experts. Of course, the management’s willingness and sound adventurism are of essence if a company is eager to pioneer the implementation of new solutions.
To what extent do market transformations impact IT companies?
Dramatically. Every company has to have an R&D center of its own. Two years ago we set up the division which now staffs more than 50 people. Moreover, these days employees need to have continuous training in parallel to their everyday work. It is the only way to fly level in the world of new technologies, because people coming from institutions of higher education can see that their hard-won knowledge is already out of date. I guess in the next 10 years we will see a cardinal change at the educational market. Online universities are already all the rage these days.
In addition, new managing methods emerge, for instance, Agile. Of course, they have some drawbacks, but they are aimed at accelerating a company’s response to external changes. They also support a customer’s early involvement into the project activities.
How is digital transformation changing personnel policies?
Personnel is the most pressing issue in IT industry. People are in increasingly short supply. Employee recruitment and retention are among the major difficulties our company is facing. Experienced specialists are not abundant, while the amount of projects at the marketplace is going up, which is a challenge for any IT company.
But in this case we can save ourselves from drowning in a swamp by pulling our own hair, as new technologies allow us to solve problems they create. For instance, in-house we implement products targeted at solving the hiring issue, and we are planning to offer such products to external customers. One of the systems, Work`n`Roll platform, generates in-house efficiency tokens and awards them to employees. In exchange for the tokens, an employee can get theatre tickets, a course voucher, a board and treatment voucher and so on. Moreover, the employees’ confidence in that they will receive bonuses in the future grows as the system supports the smart contract protocol. Additionally, employees get acquainted with new technologies, namely Blockchain, crypto-currencies, and smart contracts.
We have yet another system called iEmployee. Sometimes, we understand an employee is a bad match too late, when the employed person is leaving the company. iEmployee analyses historical data including CVs, questionnaires, candidates’ onboarding interviews and helps an HR specialist to determine if the candidate fits the bill and will stick to the company. What is more, iEmployee does not replace an HR specialist, but acts as an adviser and recommender. This centaur, the unity of a human being and an algorithm, permits acceptance of better solutions.
The importance of outsourcing is also growing. New technologies like Big Data and machine learning are very diverse, and no company can say it has expert knowledge in all innovations. So, in near future we will see more narrowly focused teams and sometimes individual experts acting as outsourcers for larger companies. Not without reason there are lively discussions of “gig economy” wh ere a huge volume of IT products will be created by integrators in collaboration with independent teams and freelancing developers.
What are the business community’s expectations from the state in connection with digitalization?
The role of decision-making authorities is very important. A while back we witnessed a complete mayhem at the crypto-currency market wrought by the lack of regulation. On the other hand, moderation is needed. Recently, three important documents on emerging technologies have been issued in Russia, the first introducing the notion of digital law and digital money, the second regulating Big Data, the third being the global program for Russia’s digital economy. Their primary virtue is drawing attention to technology innovations. As soon as the state warms to the subject, the business community has to pay attention to it too, and this is good for the market. It is not unimportant to note that a significant budget is allocated to develop the technologies.
On the other hand, the approach to producing such documents has not changed in the last 10-15 years. The main thrust is still extreme state control. For instance, the draft law on Big Data regulation places primary emphasis upon the state consolidating data about all companies supplying these data. Almost nothing is said about enabling data exchange between companies, which would give us a leap at the Big Data market.
The digital economy program sets specific quantitative targets for 2024. However, a strategic development program for IT technologies for 2012–2020 exists that will remain valid for another two years. Still, the program does not mention half of the technologies that the new digital program defines as key ones. It shows how risky it is to set quantitative targets for such a long period bearing in mind the current dynamics of IT technologies.
It would be appreciated if such documents indicated qualitative processes without setting quantitative targets that most likely will not be fulfilled. The crucial thing is that such documents should be developed with the participation of active market players dealing with these technologies and should aim in the first place at the market development to the benefit of all parties.