Run a web-based user interface
Separate songs into individual stems (vocals, drums, etc.)
Separate audio into stems using various models
