Given Monte Carlo Simulations are the base of what Range does for customers projections, making changes could be high-stakes since a tiny rounding error could throw off a retirement projection. How did you guys set up your testing to catch subtle issues?
