We call “music improvisation and accompaniment systems” those that can automatically generate melodic lines, rhythmic patterns and/or chord voicing chaining in a particular musical style. In our research, we employ different AI techniques (e.g., hidden markov models, neural networks, case-based reasoning, multi-agent systems, etc.) to successfully create systems that can work as rehearsal partners or teachers in the styles of jazz, bossa nova and other Brazilian genres. This line of research is strongly related to “sound and music analysis”, since their results may feed accompaniment systems with patterns and rules that can be used to generate music.