In the present work, we extend the classic halo model for the large-scale matter distribution including a triaxial model for the halo profiles and their alignments. In particular, we derive general expressions for the halo-matter cross correlation function. In addition, by numerical integration, we obtain instances of the cross-correlation function depending on the directions given by halo shape axes. These functions are called anisotropic cross-correlations. With the aim of comparing our theoretical results with the simulations, we compute averaged anisotropic correlations in cones with their symmetry axis along each shape direction of the centre halo. From these comparisons we characterise and quantify the alignment of dark matter haloes on the ΛCDM context by means of the presented anisotropic halo model. As our model requires multidimensional integral computation we implement a Monte Carlo method on GPU hardware which allows us to increase the precision of the results whereas it improves the performance of the computation.