The goal of AIMS is to estimate stellar parameters and credible intervals/error bars in a Bayesian manner from a set of seismic frequency data and so-called classic constraints. To achieve reliable parameter estimates and computational efficiency it searches through a grid of pre-computed models using an MCMC algorithm – interpolation within the grid of models is performed by first tessellating the grid using a Delaunay triangulation and then doing a linear barycentric interpolation on matching simplexes. Inputs for the modelling consists of individual frequencies from peak-bagging, which can be complemented with classic spectroscopic constraints. AIMS is mostly written in Python with a modular structure to facilitate contributions from the community. Only a few computationally intensive parts have been rewritten in Fortran in order to speed up calculations.