The integration of orthogonality is a crucial research strategy in the rational design of biological systems for novel applications. In order to produce such highly predictable orthogonal systems, we propose to employ cell-free systems of highly engineered composition generated from living cells, which are complex enough to reproduce the major synthetic capabilities of living cells - such as the synthesis of natural and artificial saccharides - but are simplified enough to come close to truly engineerable systems. We propose a two-step-strategy: In a first step, the protein synthesis of a growing bacterial cell will be channeled solely to a limited set of system components with the help of the RNA-interferase MazF. In a second step, the cells will be homogenized and the resulting cell-free extract, already enriched in the required protein components, will be subjected to selective hydrolysis of predefined proteins, which otherwise would connect the designed system to the remaining protein background, which would make the performance of the complex system unpredictable. As a proof-of-concept, we propose to implement a preparative 12-step synthesis from cheap glucose and N-acetyl-glucosamine to a valuable antiviral-precursor, N-acetyl-neuraminic acid.