The main hurdle in the calculation of the kinetics and thermochemistry of chemical reactions in supramolecular systems is the correct description of the molecular orbitals of the supermolecule. The errors introduced by an insufficiently populated basis set - called basis-set incompleteness errors - are quite substantial and, in the case of non-covalent intearctions, they can exceed 100% of the total binding energy. A few years ago, we introduced basis-set incompleteness potentials (BSIPs) as a way to mitigate basis-set incompleteness in density-functional theory (DFT) calculations. BSIPs are similar to effective core potentials but they do not replace any electrons. Instead, they introduce an energy correction that approximately accounts for the shortcomings of the basis set, allowing running an almost-complete-basis-set calculation at the cost of a calculation using a small or minimal basis set. In this talk, we report on the progress regarding the development of BSIPs. Specifically, a new fitting procedure based on a massive regularized linear least-squares technique is employed and we demonnstrate the excellent performance of the reulting BSIPs with a few practical examples.