This paper presents staRt, an iOS and web application for visual-acoustic biofeedback aiming to facilitate widespread uptake of technology-enhanced speech training. staRt provides a real-time linear predictive coding (LPC) spectrum that can be compared to an adjustable slider representing a target frequency for vocal tract resonance. The software currently presents exercises for children with speech sound disorder who have difficulty pronouncing "r" sounds in English, but it shows promise for adaptation to a number of other contexts.