Aleksandr Nagaev

1) Data Analysis & Modeling (Python, Pandas, SQL)

Developed a fraud detection model that increased detection of high-impact fraudulent cases by 30% (financial value) and 300% in number of cases, through micro-fraud pattern recognition.

Built a model to identify undervalued clients, reducing their rejection rate from 45% to 30%, improving customer retention and revenue potential.

2) Automation & Scalability (Airflow, ClickHouse, Python, Docker)

Automated 80% of reporting workflows and basic fraud detection for two departments, enabling near real-time monitoring and decision-making.

3) Designed and deployed ETL pipelines for large-scale time-series and array-based datasets in ClickHouse.

4) Engineering & MLOps (Airflow, Git, CI/CD, Docker)

Built and maintained production-grade data workflows and machine learning pipelines, from training to automated deployment.

Managed CI/CD for data services, ensuring rapid iteration without sacrificing stability.

5) Data Visualization & Communication (Tableau, Looker)

Created clear, actionable dashboards for both technical and non-technical stakeholders.

Presented analytical insights to executives and cross-functional teams, driving strategic decisions.

6) Collaboration & Agile Delivery (Agile, Jira, Confluence)

Worked in agile teams within matrix structures, collaborating closely with product, risk, and engineering teams to deliver on business goals.

Data Analyst — Process Automation & Scoring Systems

1) Process Automation: Designed and implemented automated workflows in Apache Airflow, streamlining complex data processing pipelines and reducing manual effort across multiple business units.

2) Bot Development & Reporting Automation: Built and maintained Telegram bot solutions integrated via REST API to deliver automated reports and alerts to business teams, enabling faster reaction to operational changes.

3) Credit Scoring Models: Developed and deployed credit scoring algorithms to assess client risk profiles, leveraging statistical modeling and machine learning techniques to improve decision accuracy and reduce default rates.

Tech Stack: Airflow, Python, REST API, Pandas, SQL, ClickHouse, Docker.

Business Impact: Increased operational efficiency by automating high-frequency reporting tasks, improved credit risk evaluation, and enabled near real-time communication of key performance metrics to stakeholders.

1) Data Visualization & BI Development: Designed and implemented comprehensive Tableau dashboards integrating multiple data sources, including ClickHouse, Vertica, and Python-based ETL pipelines. Dashboards featured multi-level groupings, dynamic filters, and key business metrics tailored for different departments and management levels.

2) ETL & Data Integration: Built and maintained robust ETL workflows using Python to extract, transform, and load large datasets from heterogeneous sources into ClickHouse and Vertica for high-performance analytics.

3) Product Analytics & Decision Support: Partnered with product teams to define key metrics, track performance, and analyze trends. Delivered impactful insights that directly influenced product strategy and operational priorities.

4) Business Impact: Helped establish a formal data-driven decision-making process within the company, improving alignment between product, marketing, and operations.