The task was towork on a dataset and build a simple text classification pipeline, starting with a baseline model and then improving it using a Transformer, followed by a minimal deployment and basic ethical reflection.