I explained that supervised learning relies on labelled data to learn a mapping from inputs to outputs, while unsupervised learning finds structure in unlabelled data (e.g. clustering, dimensionality reduction). I then walked through a supervised project where I built a model for solar energy prediction: framing the problem, collecting and cleaning weather and irradiance data, engineering temporal features, comparing gradient boosting and an LSTM baseline, tuning hyperparameters with cross-validation, and evaluating with RMSE and MAE. I also discussed how I deployed the model and monitored drift over time.