Dynamic soaring extracts energy from naturally occurring wind gradients that can be used to extend aircraft endurance, particularly in small UAVs. Autonomous thermal soaring has already been validated in flight tests with small UAVs, but this level of demonstration has not been performed for dynamic soaring UAVs, partly due to the precise wind measurements required. This paper explores wind vector sensing using flush pressure ports located on the nose of small sailplanes. Single Hidden Layer neural networks are used to generate regression models to calculate the airspeed, angle of attack, and angle of sideslip and are trained with pressure measurements collected during wind tunnel tests. These models are then applied to pressure measurements taken during flight tests of the aircraft and sensing system to judge the viability of the method.