The design of additive manufacturing processes, especially for series production in industrial practice, is of great importance for the spread of additive manufacturing technology. Manual redesign of the part’s geometry based on discrete measurement data or numerical meshes are error-prone and difficult to automate. Multiple iterations are often required to achieve the required final part accuracy. To solve these problems, a data-driven approach for geometric compensation is presented that adapts concepts from forming technology. Measurement information from an initial calibration cycle of produced parts forms the basis of the approach. Nonlinear transformations of the part geometry are used to systematically derive a new shape for the subsequent additive manufacturing process to minimize geometry deviations. Based on the purely geometric approach, the deterministic part of the components’ deviations can be compensated. The presentation covers the basic approach and analyzes its potential and challenges in series production with respect to batch fluctuations.