Conducting an analysis of the client base, building machine learning models (ML), including: - Customer segmentation; - Forecasting a tendency to buy/outflow; - Assessment of the vital value of the client (LTV). - Development and improvement of advisory systems. - Conducting a deep analysis of data and identifying key insights to support managerial decisions. - Interaction with business customers to form approaches to solving problems using ML tools, including fully developing ML services (end-to-end). - The construction of predictive models and their introduction into the production environment. - Monitoring the effectiveness of models, their training, and adjustment.

Scoring Model Development and Maintenance: - Design, validate, and sustain new application and behavioral scoring models. - Regular monitoring and analysis are performed to assess the effectiveness of the scoring model. - Propose modifications to scorecards and review validation tests. - Evaluate the financial impact of scoring model adjustments. - Calibrate scoring models as necessary and maintain internal documentation. - Implement scorecards and evaluate the relevance of external fraud-prevention services and scoring models. - Utilize web service (OpenScoring) to deploy and use it in production automatically with a daily load of 50,000 applications - Analyze and continuously monitor the effectiveness of current scoring models and decision-making tools, both internal and external. - Develop models using alternative data sources such as transaction and behavioral data. Statistical Reports and Data Visualization: - Generate comprehensive statistical reports and create data visualizations for the management team. - Conduct thorough evaluations of credit risk and anti-fraud decision-making protocols in loan origination, covering the entire product portfolio (PDL, Installment). - Optimize decision rules by leveraging best practices from global experts. - Enhance Accounts Receivable (AR) while maintaining loan quality. - Develop Risk-Based Pricing (RBP) models and contribute to constructing the risk data warehouse. - Automate dynamic reporting processes utilizing SQL Server and Power BI. - Apply data mining tools and algorithms to refine loan origination processes. Data Warehouse and ETL/ELT Processes: - Establish and manage the risk data warehouse. - Oversee ETL (Extract, Transform, Load) and ELT (Extract, Load, Transform) processes.

Preparation of statistical reports and data visualizations for the management board. A detailed review of credit risk and anti-fraud decision rules used in loan origination processes for complete product portfolio (PDL, Installment), optimizing it according to world best practices. Increasing AR with keeping the quality of booked loans. Building RBP models. Participate in building risk DWH. Automatization of dynamic reports, using SQL Server with Power BI. Using data mining tools and algorithms to loan origination processes. Development, validation, and maintenance of new scoring models (applicational and behavioral). Regular monitoring and analysis of the effectiveness of the scoring model. Making proposals for changes to the scorecards. Review of the validation tests. Assessment of the results of the changes in the financials of the business. Calibrations of the scoring models when needed. Internal models’ documentation. Deployment of scorecards; Analysis of the applicability of external fraud-prevention services and scoring models in the business; Analysis and regular monitoring of existing scoring models and external/internal tools used in the decision-making process. Analysis of their effectiveness. Creating models with alternative data, such as transaction data, behavioral data, etc. Creating Risk DWH. ETL/ELT processes.

Lectures on key subjects of IT domain, such as DBMS, programming, algorithms, machine learning, statistics, etc.

Preparation of statistical reports and data visualizations for management board. A detailed review of credit risk and anti-fraud decision rules used in loan origination processes for a complete product portfolio (SPL, Installment), optimizing it according to world best practices. Increasing AR with keeping the quality of booked loans. Building RBP models. Participate in building risk DWH. Automatization of dynamic reports, using MS SQL Server with Power BI. Using data mining tools and algorithms to loan origination processes. Development, validation, and maintenance of new scoring models (applicational and behavioral). Regular monitoring and analysis of the effectiveness of the scoring model. Making proposals for changes to the scorecards. Review of the validation tests. Assessment of the results of the changes in financials of the business. Calibrations of the scoring models when needed. Internal models’ documentation. Deployment of scorecards; Analysis of the applicability of external fraud-prevention services and scoring models in the business; Analysis and regular monitoring of existing scoring models and external/internal tools used in the decision-making process. Analysis of their effectiveness. Creating models with alternative data, such as transaction data, behavioral data, etc. All models were created with parametrical (regression models) and non-parametrical (Random Forest, XGBoost) algorithms. Creating Risk DWH. ETL/ELT processes. Assisting in creating MIS needed for risk reporting, development, and maintenance of scoring models; Liaison with members of international teams; Assistance in DWH and MIS development.

Development, validation, and maintenance of new scoring models (applicational and behavioral). Regular monitoring and analysis of the effectiveness of the scoring model. Making proposals for changes to the scorecards. Review of the validation tests. Assessment of the results of the changes on financials of the business. Calibrations of the scoring models when needed. Internal models’ documentation. Deployment of scorecards; Analysis of the applicability of external fraud-revention services and scoring models in the business; Analysis and regular monitoring of existing scoring models and external/internal tools used in the decision-making process. Analysis of their effectiveness. Assisting in creating MIS needed for risk reporting, development and maintenance of scoring models; Liaison with members of international teams; Assistance in DWH and MIS development.

Whole process of building different types (application, behavioral, anti-fraud) of scorecards. Built scorecards using blackbox tools like FICO's Model Builder and hard coding tool like R (with detailed documentation with R Markdown or R Notebooks). All kind of statistical reports and data visualizations for management board. Detailed review of credit risk and anti-fraud decision rules used in loan origination processes for complete product portfolio ( POS, car loans, cash loans, credit cards), optimising it according world best practises (worked with FICO, KPMG, AdAstra). Increasing AR with keeping quality of booked loans. Building RBP models. Participate in building risk DWH. Automatisation of dynamic reports, using MS SQL Server with MS Power Pivots. Using data mining tools and algorithms to loan origination processes. Segmentation of clients

Master thesis research on variability of geological, geochemical and geophysical data. During the internship, created database with all required data to calculate ore deposits. To calculate ore deposits, used a lot of interpolation methods, and most useful was Kriging method. All numerical calculations made in R.

Research and analyze all data concerning retail loans. Daily, weekly and monthly reports. Determining the business needs and data requirements through consultations, document analysis, surveys, site visits, business process descriptions, and task and workflow analysis. Evaluate the information gathered and distinguish between individual wants and actual business needs.