Simulations at the atomic scale reveal mechanisms and dynamical properties of microscopic processes in disciplines such as biology, chemistry and materials science. To make these simulations feasible and informative, approximate models and ingenious algorithms must be combined with the enormous power of modern supercomputers. The great success of molecular simulations is due to a rich history of innovative models and algorithms that combine low computational cost with high accuracy. Molecular simulations are already accelerating solutions to societal challenges, such as renewable energy and a circular economy, a trend that will continue in the future. My research focuses on improving the models and algorithms at the basis of these simulations to create more predictive computational models that fit into collaborations with experimental research on materials, catalysts and biological systems.