Polymath employs machine learning to transform any music library (for example, from a hard drive or YouTube) into a music production sample library. The tool automatically divides songs into stems (beats, bass, etc. ), quantizes them to the same tempo and beat-grid (e.g. 120bpm), and analyzes musical structure (e.g. verse, chorus, etc.), key (e.g. C4, E3, etc.), and other information (timbre, loudness, etc.). The end result is a searchable sample library that helps music producers, DJs, and ML audio developers streamline their workflow.