Recent computational advances in modelling of crystalline microstructures have enabled deeper insights into the microstructural evolution that governs plastic material behavior [1]. Continuum dislocation dynamics (CDD) modelling, in particular, has emerged as a powerful tool to model plastic deformation by simulating the underlying microstructural processes. However, the challenge of accurately linking simulation results with experimental data persists, especially when identifying the correct microstructural inputs that influence the mechanical response. Thus, achieving a fast and accurate derivation of simulation input parameters by translating experimental findings into a material model remains a significant challenge. The consideration of uncertainties inherent in simulation and experimental data yields another challenge that has to be addressed. In this contribution, we introduce a data-driven surrogate modelling approach to address these challenges by efficiently bridging the gap between experimental data and CDD simulations [2]. The surrogate model facilitates both forward predictions, which predict the macroscopic material response from simulation input parameters, and inverse predictions, which derive the simulation input parameter space from experimental data. We demonstrate that this approach reduces the computational costs of extensive parameter studies significantly, making it feasible to quickly explore complex microstructure-based features. By bridging the gap between simulation and experiment, this framework enhances the efficiency of material design and multiscale optimization, particularly in the context of accurate and predictive digital twins. [1] E. van der Giessen et al., Roadmap on multiscale materials modeling, Modelling and Simulation in Materials Science and Engineering 28, 43001, 2020. [2] B. Katzer et al., Combining simulation and experimental data via surrogate modelling of continuum dislocation dynamics simulations, Modelling and Simulation in Materials Science and Engineering 32, 55026, 2024.