Dinar Mingaliev Email & Phone Number

Vienna, Austria

- Improve customers' churn model in R. ROC-AUC increased from 68% to 78% (main drivers are the right choice of target balancing, feature engineering, feature selection and replacement of the logistic model by XGBoost).- Implement end-to-end churn model automation: data ingestion, data preparation, data cleaning and preprocessing, train models, model selection, prediction.- Create end-to-end telco network KPI calculation automation. The KPIs are complex Hive queries that process terabytes telco-related data (demographic, usage, geo) and associate with network performance and user experience. The application runs on a daily and monthly basis.- Productionize of MVP for network anomaly detection. R code refactoring and debugging, optimize Spark memory settings, create a unified YAML config file and make the application run on a daily and weekly basis with zero failers and warnings.- Prepare ad hoc analysis for stakeholders and visualization results in Tableau: customer movements, roamer analysis, dwell time.- DevOps. Maintain running applications on the production environment, bug hot-fix, create automation build plans in Bamboo, tickets in Jira, update the documentation in Confluence and version control in Bitbucket.Tech stack: R, Hive, SQL, Spark R, Sqoop, Tableau, Bash, Git.

Schwarzenbergplatz 3, 1010 Vienna, Austria

- Lead full-cycle credit scoring model development for a retail portfolio including data exploration, design, estimation, validation, implementation, testing, and documentation in R.- Maintain and put models into production in an internally developed framework for the whole Group- Implement automated tests for model performance monitoring in R. The main metrics are ROC AUC and accuracy.- Organize and lead a company-wide validation process for 7 subsidiaries in Europe.

Moscow, Russian Federation

- Analyze credit risk of loan portfolios (Non-Performing Loans, vintages, collections) and companies (solvency and liquidity ratios).- Prepare presentations with risk indicators and communicate at Risk Committee- Estimation of Loan Portfolio Risk based on Markov Chain Model- Lead Economic Capital Stress testing and communicate high-risk segments to senior management - Build autoregressive models (ARIMA, SARIMAX) to forecast monthly macroeconomic indicators by use of the 'statsmodels' library in Python.

Moscow, 117418, Ul. Novocheremushkinskaya, D. 69

- Calculate and monitor Risk Appetite and Capital Requirements PD, LGD (following Basel requirements) in MS SQL and VBA- Evaluate of collateral for mortgage loans in the MBS and mortgage rates in MS SQL.- Prepare portfolio risk-reports and communicate to management for decision-making.

Moscow, Russian Federation

The Manager-trainee program consists of rotations between three departments: Market research, Marketing and Sales.I. Market Research Department.- Take part in conducting focus groups to make data-driven decisions on product design, marketing campaign and communication channels. Responsible for identifying the target audience for a moss-market segment, analysis of products of competitors and advertising materials.- Collect statistical data and interpret A/B tests to assess the efficiency of advertisement and product design.- Prepare presentations with recommendations to the marketing department.II. Marketing “L`Oreal Paris” Department, hair-care (brand manager assistant).- Analyze market trends, competitors and weekly sales.- Adapt the development of international brands in Russia- Develop promotional materials and designs, organize tenders and lead end-to-end product design- Prepare presentations and communicate market insights for Head of Marketing and SalesIII. Commerce Department.- Build relationships with clients - mostly retail outlets and beauty store chains.- Negotiate the increase in the presence of the brands in the shelf space in retail outlets.- Stimulate the growth of sales in each outlet and new product launch.

Master Of Science (Ms), Management And Applied Mathematics

Activities and Societies: Leader of Institute Team in Armwrestling Courses taught: Theory of Probability, Statistics, Econometrics.

Nanodegree, Computer Vision

Upon completion of the program, I demonstrated my skills by building several models: • a facial keypoint detection neural network - takes.

Nanodegree, Deep Learning

• Convolutional Neural Networks • Recurrent Neural Networks • Generative Adversarial Networks Upon completion of the program, I.

Master’S Degree, Econometrics, Banking, Corporate, Finance, And Derivatives

High School Diploma, Physics And Mathematics, Magna Cum Laude

Activities and Societies: Leader of the School footbal team • Scored 100 (out of 100) on State Examination in Mathematics • Won the 2nd.