We present Fuzzy Prophet, a probabilistic database tool for constructing, simulating and analyzing business scenarios with uncertain data. Fuzzy Prophet takes externally deﬁned probability distribution (so called VG-Functions) and a declarative description of a target scenario, and performs Monte Carlo simulation to compute probability distribution of the scenario’s outcomes. In addition, Fuzzy Prophet supports parameter optimization,where probabilistic models are parameterized and a large parameter space must be explored to ﬁnd parameters that optimize or achieve a desired goal. Fuzzy Prophet’s key innovation is to use ﬁngerprints that can identify parameter values producing correlated outputs of a userprovided stochastic function and to reuse computations across such values. Fingerprints signiﬁcantly expedite the process of parameter exploration in ofﬂine optimization and interactive what-if exploration tasks.