At Vochlea, we have built vocal-controlled music-making software that employs real-time audio and real-time machine learning. We have used QML as it allows us to rapidly develop beautiful, hardware-accelerated user interfaces.
C++ with JUCE dominates the world of music-making software development. However, whilst JUCE provides UI functionality, it is far less productive than using QML. Qt does not have real-time, low latency audio support, therefore we have taken the step of combining the use of both libraries to get the best of both worlds. Real-time audio programming comes with important threading and memory allocation restrictions; we consider these in our architecture.