Modern organizations run on data. But “balancing” privacy with data driven innovation is not an easy task for data engineers, privacy folks and data governance experts alike. This gap leads to a very real “Cost of Privacy“ to organizations beyond just legal costs. While "Privacy by Design" has been proposed in the 1990's, it has not made its way at scale to data systems. In this paper, published in the respected IEEE bulletin on Data Engineering, the STRM team explores the opportunity to embed privacy by design directly inside data systems, and we propose a general purpose architecture to achieve it.