A network-connected predictive modeling computer comprising at least a memory and a processor and further comprising programmable instructions stored in the memory and operating on the processor, the instructions adapted to a system for predicting events, outcomes, locations, items and more, across a broad variety of applications. A preferred embodiment of the invention interfaces to a plurality of user devices associated to a plurality of Subject Matter Experts (SME) as input to compute automatic predictive intelligence models. In a preferred embodiment of the invention a combination of subject matter expertise and insight are combined by a plurality of computing devices with raw or machine-learned data to create a predictive intelligence model that exceeds accuracy and confidence in results that may be achieved by individual participants or by computation alone.