Increasing competition in supply chains forces companies into making better and computationally more demanding decisions. This becomes a bigger challenge especially in real-time applications where instances become available just before we need to start implementing the solution. Also, with the growing size of the problem instances, finding good solutions in reasonable times can be difficult even for heuristic methods. Motivated by these, we focus on variants of real-time Vehicle Routing Problem (VRP) with an additional constraint on the total time spent during the computation and loading. To lessen the computational burden, we propose a computation-implementation parallelization approach (CIP) which embeds the computation time into loading. We show that, using this approach, we can either find similar-quality solutions using less computation-only time or compute better solutions without changing the order cutoff time or the truck dispatching time. This is joint work with Prof. Joel Sokol.