Variogram modelling is a critical component of resource modelling, it provides an insight into the spatial continuity of a deposit as well as being a necessary component of all Kriging-based algorithms. Although important, it is a tedious, laborious, and difficult task that requires specialised knowledge from the modellers. This paper will present a novel approach and software to variogram modelling that is constructed around the automatic fitting. Auto-fitting is an optimisation process in which a variogram model that matches the data is found algorithmically. Instead of adding the autofitting algorithm to the traditional modelling workflow, our approach redesigns the entire modelling workflow around the autofitting requirements. First, the application can work with several domains and grade attributes at once, it efficiently computes the experimental variograms in hundreds of directions for each combination of domains and attributes. These experimental variograms are then passed to an optimisation process, the auto-fitting, that will search for the anisotropy, the ranges, the contributions, and the nugget of the 3D model. Once completed, the models are then presented in a dashboard where each model has a fitting error associated with it. The modeller can then quickly review each model and manually refine any models that are deemed unsatisfactory through a rich and interactive user interface. Through that visualisation, the modeller can inspect the model fit quality in hundreds of directions to ensure that the model properly captures the anisotropy of the data. The application also allows the updating of existing models when new data is added to a project. A modeller can then quickly update all the variogram models whenever new data becomes available ensuring that the resource model reflects the sampled data. Finally, the application is completely parallelised and can be run on the desktop or natively on the cloud.