MuSA is a web-based TEL tool developed for a Master's thesis at Universitat Pompeu Fabra. It supports the reflection phase of music practice by using saliency analysis to algorithmically identify salient moments in recorded performances based on variability musical features, offering flexible, non-prescriptive visual feedback.