ChatGPT THIS!

AI startup AudioShake—whose B2B deep-learning technology deconstructs audio into stems that can be made available for new uses in music, film, dubbing, transcription, voice synthesis and more—has announced a $2.7m seed round led by Indicator Ventures, with participation from Precursor Ventures and Side Door Ventures.

The round includes investments from musicians, songwriters, managers and music-industry executives, including Metallica-backed Black Squirrel Partners, AJR, Google’s Black Angel Group, peermusic, Audius CEO Roneil Rumburg, S-Curve Records CEO Steve Greenberg and Crush Music's Crush Ventures. The infusion of cash is earmarked for growing the company’s research and engineering teams and AI model training.

Since AudioShake’s 2020 launch, its customer list has expanded to encompass departments from all three major label groups as well as such major publishers as Primary Wave, Hipgnosis, Spirit, peermusic, Concord, Downtown and Reservoir. Hollywood film and TV studios, as well as dubbing and transcription companies, are also among its clientele.

Said AudioShake co-founder/CEO Jessica Powell, "We started AudioShake to open up audio to a broader range of experiences, making music, film and speech more interactive, editable and customizable–all while helping artists and creators expand opportunities for their work. We’ve been thrilled by the adoption of our technology across the entertainment and tech industries and are excited to now see third-party implementations of our music and speech separation technologies taking off thanks to our APIs.” The use cases for sound separation in the music and film/TV industries are vast.

AudioShake is working with labels, publishers and other rightsholders to open their recordings to new uses. Examples include separating an old VHS tape of Ol’ Dirty Bastard for sampling on SZA’s “Forgiveless” to helping fans play guitar with Green Day on TikTok.

With this recent round of funding, AudioShake has raised a total of $5m. We're told that amazing strides have been made with an AI radio-promotion engine that has 967 ways to say "Hey, Buddy."