Nikita Ryzhov

- Participated in the BASEL project for the corporate risk department as part of the data product development team for risk management and advanced analytics, aimed at improving the quality of banking risk management when working with legal entities. - Led customer-driven projects, including gathering and analysing requirements, development, testing, deployment to production, and direct communication with the customer. - Completed comprehensive training programmes in effective communication, leadership fundamentals, and situational management during employment. Key responsibilities: - Developed data pipeline workflows to create data marts from multiple sources on Hadoop, using Apache Oozie as an orchestrator and PySpark as an ETL tool. - Utilised Apache Airflow orchestrator to configure data replication from MSSQL, Oracle, and MySQL databases into HDFS storage. - Performed ad-hoc data extractions, parsing, and analysis of historical data, including parsing JSON files using libraries such as Pandas, PySpark, and others. - Created and maintained optimal architectures for data streams and data marts and their monitoring, as well as their quality control and consistency. All work was carried out in support of reporting and dashboard development for the Central Bank and risk analytics. As part of the role, developed a file generator to automate the creation of Alfa Airflow workflows. This solution increased the speed of replication workflow development by 500%, significantly improving data delivery efficiency to the end customer—the corporate risk department—for Central Bank reporting and risk analytics dashboards.

- Development of ETL processes for data marts and ML models, advanced data analytics (to identify anomalies, outliers, and inaccuracies in data) based on Hadoop and Spark. - Conducting advanced analytics (A/B testing, statistics, as well as matching) for business card sales campaigns targeting SMB (small and medium-sized businesses): increasing POS turnover, activation/reactivation of business cards, analysis of turnover and transactions. - Research into transactional behaviour patterns of business clients, formulation and validation of hypotheses to identify new intervention points in sales campaigns. Within my role, I participated in campaigns aimed at increasing POS turnover and card activation. My responsibilities included: preparing client samples, determining group sizes based on expected impact, splitting into homogeneous control and target groups, and evaluating campaign outcomes by measuring impact. As a result of these campaigns, the maximum achieved impact was: a 5% increase in average POS turnover and a 3% increase in conversion (client activation).

Contractor work (Data Engineer): 1) Designing ETL processes using Airflow (parsing and web scraping data from online sources) 2) Designing ETL processes using Apache Ni-Fi 3) Designing ETL processes using Pentaho DI orchestrated with Airflow (triggers, sensors, etc.). Data sources included Hadoop (Hive) and Vertica 4) Creating and supporting optimal data architectures that ensure efficient storage, processing, and access to large volumes of data while ensuring high quality and consistency. This includes developing optimal pipelines and data marts and monitoring them. 5) Experience communicating with business stakeholders to gather and clarify requirements for development, as well as writing access requests (Jira Service Management requests) to obtain necessary permissions In this role, I participated in the Postoffice Technologies project to migrate from the Hadoop platform to Vertica. Specifically, I developed Pentaho DI pipelines to extract data from Vertica (previously, Apache Sqoop orchestrated by Apache Oozie was used on Hadoop). These pipelines were used to build data marts for dashboards tracking postal shipments, enabling faster (up to 100%) and more reliable (with improved data flow monitoring) delivery of data for business reporting. I also contributed to the DEB (Digital Efficiency Board) project at SIBUR Digital. Within this project, I was responsible for collecting and centralising data from various sources (Hadoop, MySQL, Oracle, web APIs, etc.) into a Postgres database using Apache NiFi pipelines. This centralised data served as the source for the CDE platform (a web application), enabling efficient shift briefings and debriefings for shift supervisors on production sites.

- Разработка баз данных Oracle; проектирование структур баз данных; - Написание продвинутых SQL запросов, представлений (views), триггеров, планировщиков задач, хранимых процедур и функций на PL/SQL, а также их оптимизация; - Администрирование баз данных. В рамках рабочих задач был оптимизирован Legacy код процедур, функций, представлений (views) и триггеров в Oracle в функционале внутренней платформы для отдела кадров, учебного процесса (учебные планы, ведомости и отчеты). В результате скорость выполнения запросов была увеличена от 50% до 500% и, соответственно, был улучшен пользовательский опыт