Junior Data Scientist | Engineer
Analyse data and build tools to help review and enhance sales forecasts for retailers
Working in an agile environment
Checking for batch failures and validating the data received from the client
Generating monthly forecast accuracy reports and updating the scorecard with the received data
Reviewing and adjusting forecasts according to customer demand expectations which are based on business facts and other exterior factors (ie. weather, COVID)
User acceptance testing (UAT) and system testing
Implementing enhancements using Python for demand planning tasks automation
Performed an Exploratory Data Analysis - EDA and built an exponential smoothing model to forecast customer demand and calculate the bias before outlier detection
Pre processed the data by smoothing out promotional sales (demoting sales) and deseasonalizing the sales time series
Detected outliers using different methods : kmeans, Isolation forest, Local outlier factor and other statistical metrics like quantiles and standard deviation
Treated outliers using different metrics like the mean, the median and linear interpolation method to forecast customer demand and compare forecast bias pre and post detection
Automated the process by building an end to end pipeline that outputs a data set having the demoted sales, the deseasonalized sales, outlier indicators, and the treated outlier and visualized all results in a dashboard
Working in an agile environment
Checking for batch failures and validating the data received from the client
Generating monthly forecast accuracy reports and updating the scorecard with the received data
Reviewing and adjusting forecasts according to customer demand expectations which are based on business facts and other exterior factors (ie. weather, COVID)
User acceptance testing (UAT) and system testing
Implementing enhancements using Python for demand planning tasks automation
Performed an Exploratory Data Analysis - EDA and built an exponential smoothing model to forecast customer demand and calculate the bias before outlier detection
Pre processed the data by smoothing out promotional sales (demoting sales) and deseasonalizing the sales time series
Detected outliers using different methods : kmeans, Isolation forest, Local outlier factor and other statistical metrics like quantiles and standard deviation
Treated outliers using different metrics like the mean, the median and linear interpolation method to forecast customer demand and compare forecast bias pre and post detection
Automated the process by building an end to end pipeline that outputs a data set having the demoted sales, the deseasonalized sales, outlier indicators, and the treated outlier and visualized all results in a dashboard
نظام التوظيف التشغيلي المعتمد على الذكاء الاصطناعي أولاً، الذي يوحّد ويؤتمت دورة التوظيف بالكامل للمؤسسات الحديثة حول العالم.
د.جوب منطقة حرة ذ.م.م. 2026 © جميع الحقوق محفوظة
