Ilnaz Ziiaziev

• Improved a demand forecasting model for a 6‑month horizon, reducing WAPE score by 11 pp (36%→25%) and driving ~2% margin uplift • Built a model evaluation and monitoring pipeline (FastAPI+Airflow+Superset): served ML metrics via API, published weekly dashboards with business impact KPIs • Developed an attribution model to estimate marketing campaign revenue impact without A/B tests, achieved 28% WAPE in backtests • Designed and executed automated A/B tests to measure the incremental uplift of churn models

• Designed and deployed 5 data marts in credit scoring with end-to-end pipeline implementation in Airflow. • Implemented data quality monitoring and alerting within Airflow workflows. • Maintained a data exchange service based on Kafka between data marts and credit scoring systems.

• Automated audit reconciliation via Python/SQL and validated audit data for 4 clients.