Hello everyone! As a team of Rosselkhozbank, we look not only at what is happening here and now, but also try to understand, even feel the future of our industry. Indeed, it is better to start preparing for the changes that may come tomorrow. If you are an IT specialist in any area - from development to Dev / TestOps or project management - and are looking at the niche of financial services and products (yes, the same fintech), feel free to read on. Perhaps you will find an interesting niche and direction for yourself, in the development of which you can participate.
Fintech: what color is the ocean?
Of all the industries, the financial industry was one of the first to become a figure - right after the telecom. It is logical, because payments (in all forms and types now already) have almost completely moved to digital, from mobile applications to Central Bank systems. Banks are rapidly gaining experience in the use of technology, digital services and products are rapidly developing, often much faster than other, more mundane areas.
“An industry made up of companies using technology and innovation to compete with traditional financial institutions such as banks and financial intermediaries. Fintech includes both numerous startups and large organizations trying to optimize financial services " - this is the definition of fintech given to us by Wikipedia.
The very concept of competition has been transformed along with the market. Banks, the very representatives of the traditional sector, are often as technologically advanced as representatives of the initially digital domains, so the struggle is no longer for money, but for speed and user experience. And sometimes there is no need to compete directly - companies have learned to interact well, replacing hostility with partnerships, contracts and integration. Many are ready to become a test platform for testing future approaches: if it works, then we buy a startup, it won't work - well, it's okay, the departure of such a product will not be a blow to the market. With this approach, there is room for new bright stars who are changing the market, coming in line with leaders, often with a hundred-year history of development.
It is important that, unlike other markets, there is a global dominant player over the landscape of fintech companies - the Central Bank: it has implemented a number of projects in this area and is constantly starting new ones. The most popular is the rapidly developing fast payment system (FPS), implemented by the Central Bank in conjunction with its state fintech corporation NSPK . Clients embraced the innovation, actively transferring money to each other by phone number. The approach is constantly being updated - for example, now a QR code is enough for translation. The tool is especially interesting for small businesses, as it eliminates the cost of POS terminals.
: ( ) — , (Big Data), , , , , . CBDC ( ) , .
“Well, number and number” - you say, banks have always had computers. Did this define banks as an interesting place to work for the active IT professional? Now everything is different, “this is different”. The digitalization of the financial sector has shifted the focus of the struggle for the client - now all fintech, and not just the elite parts of high-frequency traders, are competing for a single resource - speed. The speed of connection to the service, the speed of response to the client's request, the speed of launching new products to the market and updating the existing ones (and even, let's be honest, the quick withdrawal from the market of obsolete or not taken off tools that slow down real innovations) It is these changes that form the ever-increasing demand for the development of the IT direction in all areas of modern finance. And the demand is not only for seasoned gurus - we are ready to train and raise our specialists...
The tools of such competition are on everyone's lips. Big Data, cloud systems, automation of everyone and everything, work with open APIs, system integration, Artificial intelligence and machine learning, we will tactfully keep silent about the blockchain in this list.
And even if there are no problems with money, then the shortage of developers who can make the purchased software and hardware work is already a problem. And hiring in the state can be problematic , and the services of contractors who have the resources and expertise also become more expensive.
The active build-up of IT infrastructure by banks is bearing fruit. According to research by Deloitte (available in Russian and full version in English ), in 2020 Russia entered the top ten leading countries in world digital banking. At the same time, three Russian high-tech banks were in the top thirty banks in the category "World Champions".
Finance is a special industry, where the Central Bank brings more and more new requirements that force businesses to make changes to processes and already built automation . And this, frankly, is by no means to fix the pipelines in GitLab. Even large banks with their own development teams are forced to attract specialists or entire companies from outside in order to comply with new regulatory requirements.
Such a run generates a strong demand for specialists in the field of IT systems integration.: from the ability to administer Kafka, Ignite or RabbitMQ clusters ( we also use it ) to complex BPM, BI and ETL systems. And with such things, only employees who know how to turn knobs can no longer cope - you need to immediately lay the architecture. The principle “Architecture is more important than programming” proclaimed by the Gartner agency, as well as other recommendations of this company, were taken by many banks as a guide to action. This means, in particular, an increased demand for IT Architects, and for already held Senior - new career opportunities.
Agile also plays a strong role in the IT scene. Development should not only develop, but also do it quickly. Agile methodologies, previously used by startups and small companies, were used. After unsuccessful attempts, the realization came: it will not be easy to implement Agile, “Agile for the sake of Agile” is dead, both management principles and labor motivation of people will have to be seriously changed, banks are beginning to invest serious money in personnel. The hunt begins for specialists who already have real commercial experience of flexible methodologies, let us emphasize, real , and not a book or “listened to a course on YouTube”.
Aktualochka
A decade ago, these terms were predominantly in science fiction films, and now Artificial Intelligence (AI), Machine Learning (ML) and digital assistants (chatbots) are the main technological drivers of the banking sector and we are actively engaged not only in theory , but also in solving many practical tasks, which we are writing about , in particular, for the Habr audience. There is also a place for interesting open experiments, for example, colleagues from Sbertech are experimenting with GPT-3 models .
Other areas, alas, are unlikely to ever be open to everyone. So, the heart of banking mechanisms - scoring models for assessing the creditworthiness of borrowers, use all available data sets, including public profiles and a person's digital footprint on the Internet. And the main thing now is the AI direction - the fight against cybercrime in all forms (fraud, fraud, theft of personal data). AI is used, for example, to catch atypical emissions in transaction flows and then target specific segments of operations and customers.
Well delivered in digital banking are those AI functions that are widely used in all areas of advanced retail. The main stake is on the personalization of interactions with the consumer based on his behavioral profile. Today, the first steps are being taken here, but they also help to reduce the share of inappropriate and unpleasant offers for the client.
However, not everything is so rosy. When organizing developments in the field of AI, banks often face an HR problem: young specialists with interesting experience and competencies come to them from retail and startups, but brought up on completely different tasks. If, say, in retail 3-5% of errors in the operation of algorithms is considered an excellent result, then in the banking sector there are many tasks in which even single errors are unacceptable. Here we will soon encounter the fundamental limitations of the approach itself, because AI / ML algorithms are a priori probabilistic, which means there will be a demand for serious scientific developments in this area. Doing science and don't want to be “sold” into business? Welcome to us!
They are trying to replace natural intelligence with artificial intelligence in the field of communication with the client, first of all, using chat bots. Banks' interest in voice technologies sharpened about two years ago, and during this time the path has gone from primitive talkers to rather smart interlocutors. The pinnacle of achievement can be considered the voice robot EVE.calls... In conversations with clients, Eva manifests herself as an intelligent conversationalist with a business acumen and a pleasant, welcoming voice. Few people think to suspect her "inhuman" origin. The main areas of use of such pieces are call center automation, frontal audit of operators' compliance with scripts, analysis of the most successful communication models. However, one should not think that it is possible to automate the call center and replace the staff of operators without loss of quality simply by taking a ready-made solution. The "box" is a very rough approximation, rather a caricature of a person. Its additional training for specific tasks of a particular business is the work of very expensive specialists, whom you should also try shanty.
Banks? Open APIs ?? Are you joking?
In the fall of 2020, the first public discussions of the open banking standards proposed by the Central Bank , which are advisory in nature , took place . Open API standardization benefits most market players, since it simplifies and makes cheaper various types of integrations: integration of partner IT systems, integration of services and channels, connection to ecosystems and marketplaces. And all this, finally, using really open specifications and technologies familiar to everyone in the world of web development, for example, OpenID Connect. With the development of the concept of open banking, the demand for qualified specialists developing or promoting APIs is growing. At the same time, the transition to API standards requires significant investments from large players, especially in terms of information security. But, we hope, soon it will be as easy to integrate the functions of your bank into your service as a VK widget. Already having my own experience within the Svoe Rodnoe marketplace, we see great prospects in this direction.
“Stop, what if I deny all these newfangled AI / ML trends, since I’m a really tough administrator, DevOps or, hurray, a system programmer in C / C ++ or Java?” Don't worry, you won't be sent to retrain. For all these fancy things to work, you need serious infrastructure - databases, real-time processing systems , distributed in-memory databases, monitoring and millions of other systems. Some of this is in OpenSource, and banks are happy to use and even actively contribute to world-class projects. Banks do something themselves or with partners (the same Tarantool ), demonstrating the competitiveness of their own developments. And that's not all.
Over the horizon
We would venture to suggest that in the near future nothing will change globally, but what already exists will develop and competition will only intensify.
The Central Bank will continue to be a serious infrastructure player, up to the point of ousting commercial participants from some niches. Today the idea of creating a state platform for financial cybersecurity is being discussed. And some of the services previously provided only by banks can now be done by other players, offering add-ons over the infrastructure level of the service of fast payments from the Central Bank.
State pressure on players will increase, stimulating the transition to Russian software. But the transition will not be quick, and in some cases impossible or monstrously costly. On the one hand, most banks do not have the money for such a massive IT restructuring. Moreover, many foreign solutions have managed to take root and acquire links with other fragments of the infrastructure. And there is no open API (and often even normally documented) there. On the other hand, there are no worthy Russian analogues for some types of software. For example, many banks have been using Oracle DBMS for a long time. There are no Russian DBMSs of this level, the closest competitors in this field are PostgreSQL, Tarantool and Apache Ignite. They will need to be finalized, adapted and built into processes - so that everyone will have work.
The demand for information security specialists will grow following the growth of business investments in the cybersecurity sector. The digital criminal business is rapidly developing and increasing technological equipment, so it will not be possible to solve the information security problem once and for all. The problem of fraud is increasingly moving from the technical sphere to the field of psychology: in 90% of cases of money theft, social engineering is used. So the most realistic way to solve the problem is to direct part of the security budgets to increase the literacy of the population, and not only to new IT solutions within banks.
This is how our fintech looks like from a bird's eye view. If you have any questions or are interested in details - write. We will tell you what we know.