We can have an automatic annotation of the various parts of the pipeline if we work with structured data.