At Kodland, I redesigned the next-month sales-base forecasting methodology, improving first-month forecast accuracy from about 60% to over 80%. I use statistical analysis to evaluate A/B tests on student retention and reducing manager workload after missed lessons.
I automated teaching payroll using DWH data, dbt, Python, and dashboard-based exception validation, reducing processing time from about a week to a couple of days. I also automated internal-sales error verification, replacing manual student-by-student checks with a dashboard-generated error list.
I built sales data marts for conversion, cohort, and operational KPI analysis, and created a teaching-cost data mart to support cost-per-student analysis. I also redesigned internal-sales data marts and metric definitions, reducing refresh time from more than an hour to about 15 minutes.
Before becoming a Data Analyst at Kodland, I worked as an Administrator, automating operational processes and building Google Sheets tools for employee performance KPIs. I also developed Python scripts for recurring data-warehouse exports. In a personal project, I built and deployed the Embassy Telegram Bot for citizen registration, document processing, and embassy information.
