In this paper we present a platform that mixes a domain-specific language for stochastic discrete dynamical models implemented with LLVM and a C++ multi-threaded library for the simulation, analysis and statistical evaluation of these models. More precisely, the user can easily implement a dynamic model, specifying the state and exogenous variables, parameters, state and observation functions, noises, and then run simple simulations, sensitivity analysis, parameter estimation, data assimilation or uncertainty analysis on computation clusters. We believe that this platform can foster good modeling practices since it simplifies the management of data, models and simulations on clusters and it demonstrates the different steps for a proper model design and evaluation. This platform was initially developed for the plant growth modeling community, for which such methodological tools are deeply needed, since a large variety of models coexist in the literature with generally an absence of benchmarking between the different approaches and insufficient model evaluation. However, the software can be used in any scientific field for which discrete dynamical models are developed.