Senior Data Scientist
Eurasian Bank
A commercial bank in Kazakhstan offering retail and corporate banking, including lending, cards, payments, and deposit services.
Oct 2024 — PresentAnalyzed lending and payment behavior, ran app and campaign tests, and shifted communications to push notifications, cutting SMS costs by 600M KZT.
• Conducted in-depth loan portfolio analysis, reactivating suspended lending products, and optimizing credit limits using Python (Pandas, NumPy, Plotly).
• Performed payment behavior analysis, enabling targeted in-app payment campaigns for intra-bank transfers.
• Identified underutilized MCC codes and inactive clients, launched district-level promotional campaigns, and discovered potential private banking clients.
• Executed A/B testing for app scaling, optimizing communication channels, and reducing SMS costs by 600M KZT through push notifications.
Senior Data Scientist
Halyk Bank
Kazakhstan's largest bank by assets, serving retail and business customers across banking, payments, lending, and digital financial services.
Dec 2023 — Jul 2024Analyzed customer cash flows and built segmented marketing and automated partner campaigns, increasing MAU by 19% and boosting campaign engagement.
• Led a large-scale analysis of cash flow movements using Plotly Sankey and Polars, visualizing client behavior and enabling new product and commission strategies.
• Increased MAU by 19% through segmentation and personalized marketing, applying a Lookalike Model (LightGBM + cosine similarity) for audience expansion.
• Delivered 23 automated push campaigns for partner promotions via Apache Airflow, integrating data from Oracle, PostgreSQL, ClickHouse and using advanced SQL (CTEs, window functions) — boosting engagement and conversions.
Senior Data Scientist
Jusan Bank
A commercial bank in Kazakhstan serving retail and business customers with products such as cards, payments, lending, and deposits.
Jul 2022 — Oct 2023Built a gamified card campaign, KPI monitoring, and app recommendations, with card openings up 72.3% and manual reporting down 95%.
• Designed and implemented the gamified campaign ”A Million for Good Deeds” using FastAPI, Oracle, PostgreSQL, and Airflow, increasing new card openings by 72.3% and reactivated cards by 26.3%.
• Built a unified KPI monitoring system with Grafana and Telegram alerts, reducing manual Excel reports by 95%.
• Enhanced mobile app personalization through a recommendation engine that dynamically adapts the homepage to 100% of users, allowing fully individualized product offers.
Middle Data Scientist
Jusan Bank
A commercial bank in Kazakhstan serving retail and business customers with products such as cards, payments, lending, and deposits.
May 2021 — Jun 2022Developed bonus budget forecasts, assembled loan scoring features, and automated ingestion of customer and utility payment data.
• Developed a bonus budget prediction model using XGBoost and Random Forest, improving budget allocation accuracy.
• Created a feature store for loan scoring, aggregating demographic, financial, and behavioral data from PostgreSQL, ClickHouse, Oracle with SQL (window functions, CTEs).
• Automated data ingestion (including utility payment parsing) via Selenium, BeautifulSoup, and Apache Airflow, strengthening credit scoring reliability.
Junior Data Scientist
Jusan Bank
A commercial bank in Kazakhstan serving retail and business customers with products such as cards, payments, lending, and deposits.
Apr 2020 — Apr 2021Automated daily reporting and built real-time bank dashboards for customer activity, queues, ATMs, and cash availability.
• Automated internal reporting using Python (pandas, smtplib, openpyxl) to provide daily insight on transfers, payments, and loans to executives and product owners.
• Built real-time dashboards (MAU/DAU/WAU, queues, ATM status, cash availability) with Plotly Dash and Grafana, creating a unified data source for the bank.
