Optimizing production processes is key to increasing product qualities and quantities while reducing production costs. Moreover, due to stricter sustainability requirements, production processes must be assessed regarding a growing number of objectives (e.g., resource and energy consumption or waste production). Given these increasing complexities, traditional statistical methods, such as the design of experiments, become inefficient since the number of experiments increases exponentially with the number of factors. Bayesian optimization is an optimization method that uses the Bayesian Theorem to iteratively select new factor levels aiming to maximize the information content of the experiment. After evaluating an experiment, the model updates itself and proposes new factor level sets until the desired product or process performance is obtained. Thereby, Bayesian Optimization offers significant efficiency improvements in process design, defect containment, and continuous adaptation of production processes. The goal of the Evolve project was to develop a software application to apply Bayesian Optimization for tailor-made process optimization for two use cases: ultrashort pulse laser structuring of metallic surfaces and cultivation of plant suspension cells in a batch fermentation bioreactor. For laser structuring, the goal is to minimize the processing time, while achieving a target surface roughness. For the plant suspension cell use case, researchers aim to optimize the cultivation medium to reduce nutrient consumption while maximizing growth rate and absolute biomass yield. A special focus was on enabling multi-objective optimization and integrating Bayesian Optimization into production engineering practice. The two case studies demonstrated that our Bayesian Optimization software can be applied to different processes due to its configurability. The software provided both plausible parameter sets and experiments with high information content for the experts, which substantially reduced the number of experiments required for process optimization. Both the involvement of product and process experts in the configuration of the parameter space and the definition of the target variables as well as the graphical visualization of the results further increased the transparency and acceptance of the system by the operators. We have shown that the Bayesian Optimization outperforms the classical design of experiments in terms of configurability of the experimental design, comprehensibility of the procedure, and plausibility of the results, which is crucial for the acceptance of the technology in production engineering practice. Further research will focus on the incorporation of disturbance variables, the configuration of the Bayesian algorithm depending on process characteristics, and the specifiability of compliant process objective functions.