In this paper, we present a method to generate state-resolved reaction cross sections in analytical form for rovibrational energy excitation and dissociation of a molecular gas. The method is applied to an ab initio database for the N 2 -N system developed at NASA Ames Research Center. The detailed information on N 2 +N collisions contained in this database has been reduced by adapting a Uniform RoVibrational-Collisional bin model originally developed for rate coefficients. Using a 10-bin system as an example, a comparison is made between two sets of coarse-grain cross sections, obtained by analytical inversion and direct binning respectively. The analytical inversion approach is especially powerful, because it manages to compress the entire set of rovibrational-level-specific data from the Ames database into a much smaller set of numerical parameters, sufficient to reconstruct all coarse-grain cross sections for any particular N 2 +N-collision pair. As a result of this approach, the computational cost in in large-scale Direct Simulation Monte Carlo (DSMC) flow simulations is significantly reduced, both in terms of memory requirements and execution time. The intended application is the simulation of high-temperature gas-dynamics phenomena in shock-heated flows via the DSMC method. Such conditions are typically encountered in high-altitude, high-speed atmospheric entry, or in shock-tube experiments. Using this coarse-grain model together with ab initio rate data will enable more accurate modeling nonequilibrium phenomena, such as vibrationally-favored dissociation, an effect that is not well-captured by the conventional models prevalent in DSMC (i.e. Larsen-Borgnakke and Total Collision Energy).