Phase 1: Proof of Concept
The goal of this phase is to prove out the general concepts utilized by TensorX in a way that provides confidence that there is merit to its approach. The forecast will be generated initially based only on historic data. Subsequently, TensorX will add in volume data to improve and refine the forecasting.
Phase 2: Forecasting and Economic Modeling
A key component of the TensorX approach is the use of Economic Triage. The most important outcome of this phase is an economic model in software that not only forecasts the price and volume of particular products but also quantifies the value of the improved accuracy in the forecasts. It is our experience that once a problem can be modeled and implemented in software, using all of the rules, constraints, and variable pricing involved, then such a model can be utilized to optimize business performance.
Phase 3: Business Process Optimization
Once a simulation model is effectively constructed, the model can be utilized to optimally solve the problem in question. By generating the insights in combination with the value of the actionable intelligence, TensorX can work with the client to tune the system to answer specific questions to further its operational objectives. TensorX can also use the optimization process to run “what-if” analyses to evaluate the effect of various changes in the economic value chain. In addition to utilizing Machine Learning techniques available vis a vis a Neural Network infrastructure, TensorX often utilizes sophisticated Genetic Algorithms in the optimization of large-scale decision support.