Graphene-derived materials are of significant interest in the manufacturing of functional devices due to their exceptional physical and chemical properties. The challenge in standardizing the synthesis of pristine graphene layers makes patterning of carbon materials a more favorable fabrication method. However, the graphene patterning process is often impeded due to high temperatures or toxic chemicals. Thus, it is paramount to find a straightforward and reproducible processing technique in a large scale. Laser reduction of graphene oxide and polymers have gained attention since it is a one-step chemical-free process to fabricate graphene circuits on-the-fly down to micrometer scales. The selective and locally confined laser-reduced circuits can be written on almost any solid or flexible substrate, showing promising potential in the manufacturing of advanced devices such as supercapacitors, energy storage devices and biosensors. In our work, we investigate strain effects on the electrical properties of graphene circuits on flexible substrates. We also utilize in-situ Raman and optical analysis with a view to automate the fabrication process in a closed-feedback loop. Furthermore, we present results of state-of-the-art model-based optimization in materials science. In particular, we use surrogate models for automated parameter tuning to optimize the manufacturing of laser-reduced graphene circuits. Our results are reproducible even with sparse training data and we discuss possibilities of human-less advanced manufacturing.