June 10, 2019 at 14:13Graduate of Netology course “Data Science” about his work in the banking sector
In Netology, the direction of Data Science appeared in 2016. When we first started, there were fears: the field is new, the demand for Scientists' dates for the companies, although decent, but there was not a large flow of people wishing to enter the sphere, and there are also a lot of free English-language resources for self-training on the network, which is why we took risks.
But today, there are already 10 courses in various specializations in working with data, and the number of graduates is more than 800. We decided to ask one of such graduates about his work with data, how he came to the field, how Machine Learning develops in Loco Bank and what kind of people he is looking for in his team.
Vyacheslav Potapov, head of data analysis and machine learning areas Loko-Bank and a graduate course « the Data Scientist »:
I graduated from Moscow State Technical University. Bauman in the specialty "Spacecraft" and upper stages in 2011. After that, he worked for 7 years in various places as an analyst, database developer and warehouse architect. During this time, I learned a lot about data processing and storage, but at some point I wanted to dive more into analysis - to understand what all these numbers mean, what I store and process.
I began to look for directions for growth: I studied related positions in IT, looked at what level of salaries in the industry and what is more in demand. There were many articles on Habr and videos on Youtube, to some extent they helped me understand the essence of working with data and how my existing skills at that time could be useful.
Then I met with Data Science (DS) and Machine Learning (ML), but the fundamental foundation was not enough. The field is very wide and when you watch some videos or articles, you get only fragmentary knowledge, but in general there is no understanding what the essence of the specialty is, what are the directions, methods, tools. This is how to read a thick textbook on mathematics for universities, but without explanation and practice, it will be difficult to apply the knowledge gained.
A colleague told me about Netologiya, where there was a large full-time program in Data Science, and I did not meet such suitable offers in the Russian-speaking market. As a result, he successfully unlearned and defended his thesis on the topic “Image Recognition Using Neural Networks”. As I remember now, it was very difficult, I did not have the practice of solving full-fledged tasks, and I really wanted to do not just an educational work, but a fully working project.
In parallel with his studies, he tried to solve problems with Kaggle and to do work projects.
And right after the course I started looking for a place where I could fully engage in data analysis, since it is difficult to combine the work of the BI system architect and practice in DS.
After a series of interviews, he chose Loko-Bank and the direction of DS.
In Loko-Bank, they saw these prospects - so I began to work in the Digital Business block, which is developing the direction of analytics.
Now the bank has a classic IT department that is responsible for infrastructure and data storage, other departments use these data sources and set requirements for the integration of new ones. In total, about 40 employees work in the company with analytics.
At Loko Bank, process automation, data analysis and building a data-driven economy are becoming the company's priorities. I hope that based on the information we will be able to more correctly build sales, conduct risk assessments and the entire business.
In the business unit, work with analytics is divided into two areas: classic analytics - BI, whose specialists analyze the planned and actual indicators of the company, prepare reports on sales, balances, income and expenses and ML direction.
Machine Learning focuses on creating algorithms that make predictions based on evidence from classic analysts, generate new data and look for hidden dependencies and anomalies. This is the department I am in charge of.
ML in the bank is just starting to develop. But I have a goal - to build a system so that it helps the business and allows you to use all modern approaches to increase revenues and reduce costs. We have to completely change the business processes and look for ways to implement machine learning tools in the existing IT architecture. It can be difficult with this, since the architecture was not designed yesterday, and some of the requirements were simply not laid down in it.
For example, requirements for collecting logs for customers to enter a mobile bank. For classical analytics, they are not needed, so they were never collected or stored. I explained that on the basis of these logs, we can train the model to make predictions on the platform load and see the relationship between the use of a mobile bank and customer’s profitability. And if it were not for the development of ML, such analytics would simply not exist, because no one would deal with this issue. What was needed was a guide who would explain why and why, give directions, how to build architecture, how to collect data, how to build models, where to apply them.
With the introduction of machine learning, I want to build a culture of working with data in the bank as a whole: their collection, processing, as well as the integration of new sources. At the same time, we are already solving the tasks of predictive analytics for customers, we are engaged in their segmentation in order to then optimize tariffs and increase sales of the company.
We are also engaged in financial monitoring, we analyze suspicious clients and transactions. Now the company spends a huge amount of human and financial resources on this task. And we want to simplify and make these processes more efficient.
If we talk about what has already been done, then we started collecting and storing data, in particular user logs, about which I wrote above. Now we store information on the history of changes in the customer card in the Federal Tax Service.
At the moment, we are developing a model to determine the negative behavior of customers (legal entities and individual entrepreneurs) and have already received the first good results. Score for one of the popular metrics is 0.86. Of the algorithms we use gradient boosting. In the near future we plan to achieve stability in its work, including by connecting additional sources. This model should help reduce company risks and optimize the costs of finding dishonest customers.
Our team is only being formed, so now I try to take the generalists. Of course, a person may be more inclined to develop or, conversely, to business analysis, but nevertheless he must understand the process of creating a solution as a whole, understand his role in it. This is a good option for those who want to try themselves in different roles.
It is important that a person knows how to solve real practical problems, at least can explain the approach and set of steps. At the interviews, I try to give logic problems, and I ask for a general understanding of algorithms and techniques, without mathematics.
Creating an ML solution is far from a trivial task, so it’s not enough just to take all the data, throw it into the algorithm and wait for a miracle. You need to be able to immerse yourself in the subject area, be able to communicate, ask and listen, somewhere these skills may turn out to be even more valuable than technical ones.
More specifically, the department is now primarily interested in Big Data engineers. Neural networks and xgboosts are good, but first you need to find specialists who can collect the correct, prepared data in large quantities. Without them, no machine learning will work. I need at least two people in this direction. But the company has many requirements for them: they must know ETL tools, SQL and have experience in building storefronts and data warehouses, as well as be able to solve optimization problems.
It would also be nice to supplement the staff with two analysts, preferably with experience in the banking sector. And although Data Science is a priority, the field can be any.
To solve this problem, you need to understand the business itself and the existing tools, as well as have good soft skills to correctly present the solution to the problem. And it is extremely difficult to find such.
Since we are only just introducing ML in business companies, we need to implement a number of decisions on which further confidence in the entire area will depend. These decisions are related to the rationale for the existence of a department for business. Machine Learning is now well known to everyone, so it is of particular interest.
After the successful implementation of ML tools within my department, we plan to expand the pool of tasks and the staff of specialists throughout the bank.
On the one hand, there are customers who want to get good service and save their data, and on the other hand, there are always people who want to access storage facilities for confidential information.
And a person who wants to come in the direction of Machine Learning in the banking sector must be able to correlate ML work tasks with the main goals of the bank in the first place.
First of all, it’s worth answering yourself the question of what exactly you want to do, and only after that look what is needed for this. DS is a huge area for development, and on the one hand it’s good, but on the other hand, you can wander for a very long time and not come to something specific.
In the beginning, I would not recommend diving deep into math. Focus on solving practical problems and tools (libraries, methods). I was greatly helped by the experience of developing databases, cleaning and processing data, and initial analysis. In real work, it is data collection and preparation that occupy most of the time, and high-quality work in this direction will significantly improve the quality of ML solutions in the future.
It's great that we live in a time when any information can be easily found. The network has many courses in various fields, communities (ODS), conferences and workshops are held periodically. But you need to understand that ML is a young discipline, it is only being formed and there is no fundamental approach to learning. Therefore, development paths must be chosen carefully: to study different training programs, to set the right accents for yourself. I was lucky - I chose a course that met my requirements and expectations, and led to the development of a huge and promising direction in Loko-Bank.
But today, there are already 10 courses in various specializations in working with data, and the number of graduates is more than 800. We decided to ask one of such graduates about his work with data, how he came to the field, how Machine Learning develops in Loco Bank and what kind of people he is looking for in his team.
Vyacheslav Potapov, head of data analysis and machine learning areas Loko-Bank and a graduate course « the Data Scientist »: I graduated from Moscow State Technical University. Bauman in the specialty "Spacecraft" and upper stages in 2011. After that, he worked for 7 years in various places as an analyst, database developer and warehouse architect. During this time, I learned a lot about data processing and storage, but at some point I wanted to dive more into analysis - to understand what all these numbers mean, what I store and process.
I began to look for directions for growth: I studied related positions in IT, looked at what level of salaries in the industry and what is more in demand. There were many articles on Habr and videos on Youtube, to some extent they helped me understand the essence of working with data and how my existing skills at that time could be useful.
Then I met with Data Science (DS) and Machine Learning (ML), but the fundamental foundation was not enough. The field is very wide and when you watch some videos or articles, you get only fragmentary knowledge, but in general there is no understanding what the essence of the specialty is, what are the directions, methods, tools. This is how to read a thick textbook on mathematics for universities, but without explanation and practice, it will be difficult to apply the knowledge gained.
A colleague told me about Netologiya, where there was a large full-time program in Data Science, and I did not meet such suitable offers in the Russian-speaking market. As a result, he successfully unlearned and defended his thesis on the topic “Image Recognition Using Neural Networks”. As I remember now, it was very difficult, I did not have the practice of solving full-fledged tasks, and I really wanted to do not just an educational work, but a fully working project.
In parallel with his studies, he tried to solve problems with Kaggle and to do work projects.
And right after the course I started looking for a place where I could fully engage in data analysis, since it is difficult to combine the work of the BI system architect and practice in DS.
After a series of interviews, he chose Loko-Bank and the direction of DS.
It seems to me that Data Science, as some analogue of the research institute, needs trust, patience and understanding of perspectives from the management.
In Loko-Bank, they saw these prospects - so I began to work in the Digital Business block, which is developing the direction of analytics.
What analysts and Data Scientist are doing at Loko Bank
Now the bank has a classic IT department that is responsible for infrastructure and data storage, other departments use these data sources and set requirements for the integration of new ones. In total, about 40 employees work in the company with analytics.
At Loko Bank, process automation, data analysis and building a data-driven economy are becoming the company's priorities. I hope that based on the information we will be able to more correctly build sales, conduct risk assessments and the entire business.
In the business unit, work with analytics is divided into two areas: classic analytics - BI, whose specialists analyze the planned and actual indicators of the company, prepare reports on sales, balances, income and expenses and ML direction.
Machine Learning focuses on creating algorithms that make predictions based on evidence from classic analysts, generate new data and look for hidden dependencies and anomalies. This is the department I am in charge of.
ML in the bank is just starting to develop. But I have a goal - to build a system so that it helps the business and allows you to use all modern approaches to increase revenues and reduce costs. We have to completely change the business processes and look for ways to implement machine learning tools in the existing IT architecture. It can be difficult with this, since the architecture was not designed yesterday, and some of the requirements were simply not laid down in it.
For example, requirements for collecting logs for customers to enter a mobile bank. For classical analytics, they are not needed, so they were never collected or stored. I explained that on the basis of these logs, we can train the model to make predictions on the platform load and see the relationship between the use of a mobile bank and customer’s profitability. And if it were not for the development of ML, such analytics would simply not exist, because no one would deal with this issue. What was needed was a guide who would explain why and why, give directions, how to build architecture, how to collect data, how to build models, where to apply them.
With the introduction of machine learning, I want to build a culture of working with data in the bank as a whole: their collection, processing, as well as the integration of new sources. At the same time, we are already solving the tasks of predictive analytics for customers, we are engaged in their segmentation in order to then optimize tariffs and increase sales of the company.
We are also engaged in financial monitoring, we analyze suspicious clients and transactions. Now the company spends a huge amount of human and financial resources on this task. And we want to simplify and make these processes more efficient.
If we talk about what has already been done, then we started collecting and storing data, in particular user logs, about which I wrote above. Now we store information on the history of changes in the customer card in the Federal Tax Service.
At the moment, we are developing a model to determine the negative behavior of customers (legal entities and individual entrepreneurs) and have already received the first good results. Score for one of the popular metrics is 0.86. Of the algorithms we use gradient boosting. In the near future we plan to achieve stability in its work, including by connecting additional sources. This model should help reduce company risks and optimize the costs of finding dishonest customers.
What kind of specialists are needed for ML direction
Our team is only being formed, so now I try to take the generalists. Of course, a person may be more inclined to develop or, conversely, to business analysis, but nevertheless he must understand the process of creating a solution as a whole, understand his role in it. This is a good option for those who want to try themselves in different roles.
It is important that a person knows how to solve real practical problems, at least can explain the approach and set of steps. At the interviews, I try to give logic problems, and I ask for a general understanding of algorithms and techniques, without mathematics.
Since I myself am an engineer, I try to look for people with an engineering background in my team, although this is not a taboo. I know examples when people came into the profession without technical education.
Creating an ML solution is far from a trivial task, so it’s not enough just to take all the data, throw it into the algorithm and wait for a miracle. You need to be able to immerse yourself in the subject area, be able to communicate, ask and listen, somewhere these skills may turn out to be even more valuable than technical ones.
More specifically, the department is now primarily interested in Big Data engineers. Neural networks and xgboosts are good, but first you need to find specialists who can collect the correct, prepared data in large quantities. Without them, no machine learning will work. I need at least two people in this direction. But the company has many requirements for them: they must know ETL tools, SQL and have experience in building storefronts and data warehouses, as well as be able to solve optimization problems.
It would also be nice to supplement the staff with two analysts, preferably with experience in the banking sector. And although Data Science is a priority, the field can be any.
The main problem of the market is the lack of people who can translate the needs of the business into a meaningful ML-task, and sometimes propose a solution proactively.
To solve this problem, you need to understand the business itself and the existing tools, as well as have good soft skills to correctly present the solution to the problem. And it is extremely difficult to find such.
Where to develop
Since we are only just introducing ML in business companies, we need to implement a number of decisions on which further confidence in the entire area will depend. These decisions are related to the rationale for the existence of a department for business. Machine Learning is now well known to everyone, so it is of particular interest.
After the successful implementation of ML tools within my department, we plan to expand the pool of tasks and the staff of specialists throughout the bank.
A bank is, first of all, large flows of data, a large customer base and, accordingly, a huge responsibility.
On the one hand, there are customers who want to get good service and save their data, and on the other hand, there are always people who want to access storage facilities for confidential information.
In my opinion, with the growing workload and complexity of the processes, the delegation of some responsibilities and functions to machines is the only possible condition for the stable growth of the company.
And a person who wants to come in the direction of Machine Learning in the banking sector must be able to correlate ML work tasks with the main goals of the bank in the first place.
Tips for those who want to enter the Machine Learning field
First of all, it’s worth answering yourself the question of what exactly you want to do, and only after that look what is needed for this. DS is a huge area for development, and on the one hand it’s good, but on the other hand, you can wander for a very long time and not come to something specific.
In the beginning, I would not recommend diving deep into math. Focus on solving practical problems and tools (libraries, methods). I was greatly helped by the experience of developing databases, cleaning and processing data, and initial analysis. In real work, it is data collection and preparation that occupy most of the time, and high-quality work in this direction will significantly improve the quality of ML solutions in the future.
It's great that we live in a time when any information can be easily found. The network has many courses in various fields, communities (ODS), conferences and workshops are held periodically. But you need to understand that ML is a young discipline, it is only being formed and there is no fundamental approach to learning. Therefore, development paths must be chosen carefully: to study different training programs, to set the right accents for yourself. I was lucky - I chose a course that met my requirements and expectations, and led to the development of a huge and promising direction in Loko-Bank.
From the editors
- Mitap “ Career in Data Science for Beginners ” - August 11, Moscow + broadcast
- Profession Data Scientist , 10 months of training from 16 667 p. / month
- Courses " Machine Learning ", " Big Data ", " Python for working with data "