Igor Iakubovskii
Portfolio
Appbooster
I developed a causal inference model to optimize online promotional spending for a client's app, significantly cutting their marketing costs. I crafted this tool as a standalone application using Streamlit and Tableau. It allowed our account managers and clients to customize advertising campaign parameters and receive predictions for the optimal number of motivated app installs. This innovation ultimately reduced the expenses associated with promoting the mobile application. Technical stack for this project 1. ETL - PostgreSQL 2. Research and Development - Python 3. Production: - CI/CD Github actions - Airflow - Docker 4. Visualization - streamlit and tableau
SamokatTech
Designed and established a comprehensive pipeline for calculating elasticity coefficients for the all categories in marketplace, leading to more efficient marketing budget allocation. Successfully implemented promo calculator for our managers to predict GMV depending on parameters of potential promo campaign Technical stack for this project 1. ETL - Hadoop PySpark, polars, greenplum 2. Research and Development - Python 3. Production: - CI/CD Gitlab - Airflow - Docker 4. Visualization - google sheets, streamlit, tableau
P2P.org
Created dashboards for internal and external clients in Superset and Tableau Spearheaded the creation and maintenance of efficient ETL processes from the ground up. The bunch of dashboards, which I created, was for monitoring revenue, effectiveness of our validator and our clients Technical stack ETL: - Python for gathering data from different nodes - BigQuery Orchestration: - dbt - Airflow Production - Superset - Tableau