My experience comes from nearly five years as a hands-on, dedicated specialist leading ML data operations at Antares (AutoTune). In that time I managed vendors, worked alongside engineers, built annotation schemas, recorded and built datasets myself, and negotiated ethical legal agreements directly with artists. See some of my work here.

I started Stylus for the same reason I started building custom audio datasets at Antares: to prove that high specifications and quality standards are not only achievable, but affordable and fast to turn around. As a musician, producer, and audio engineer, I find collaborating with artists and audio professionals to build our creative toolkit deeply rewarding.