The task was to predict whether employee signed up for the program would resign in the future (in the end it turned out that we are supposed to detect which of the employees in the dataset have already resigned - in my opinion it was a significant change of the task).
My team solved the problem using Gradient Boosting and won the first prize.
Visualisation - exploratory data analysis, graphs showcasing some interactions within data
Prediction - processing the data and training XGB model