Voice interfaces have rapidly gained popularity, introduc-ing the opportunity for new ways to explore new interac-tion paradigms for music. However, most interactions withmusic in current consumer voice devices are still relativelytransactional; primarily allowing for keyword-based com-mands and basic content playback controls. They are lesslikely to contextualize content or support content discoverybeyond what users think to ask for. We present an approachto dynamically augment the voice-based music experiencewith background information using story generation tech-niques. Our findings indicate that augmentation can havepositive effects on voice-based music experiences, giventhe right user context and mindset.