The current state of aircraft model validation has not change dramatically in several decades. Stabilized test points and unique flight test maneuvers are still utilized to determine a platforms acceptability as compared to a model. A model that has be developed from some analysis but mostly anecdotally from previous test programs. This in conjunction with the assumptions and simplifications used in post-flight processing the data has resulted in test crews having to attempt to recover from unanticipated and undetected problems during flights. Additionally, new model development is not readily performed as the programmatic impact is not typically acceptable. Our solution is to utilize existing processing power and networking capabilities and leverage modern statistical and machine learning techniques to provide dynamic, near real-time model solutions during flight. In parallel this approach will develop novel models based on the live data. Innovative aircraft designs can benefit from this as well since the initial model selection, which is difficult, is less critical and acceptable model fidelity will be developed faster and in flight. Significant efficiencies will also be obtained as a greater amount of the flight time and test data will be utilized, which will provide improved confidence levels for resultant analysis.