Each notebook in this repository looks at an AB experiment dataset and exemplifies how onn can go about exploring, cleaning, and describing the data first; and applying statistical tests to come to a conclusion (or not) about the AB testing experiment results.
Each experiment is different and so the data treatment, while having some similarities, is always tailored to each dataset.
Most of the datasets I've obtained from Kaggle. At the beginning of each notebook there will be a link to the source of the dataset, or an explanation about where it came from.