Data

1. Real life examples?
2. Anonymisation of existing data in the various environments if applicable?
3. Do we need to baseline the data before deployment for comparison?