Design a data transformation pipeline that can take raw data through a series of transformations. The pipeline should allow chaining of transformation functions, support for parallel transformations, logging of transformation steps, and error handling with the ability to continue past certain errors or stop on critical failures.