AudioShake: using AI to take control of audio content

This week, Yanni and Andy chat with Suzanne Kirkland about how the company uses AI to separate audio into its “stems” or component parts, giving organisations more control over their audio and video assets. In sports, that means tweaking levels or replacing unwanted sounds to make content libraries more usable. But the tech, which was born out of a love for homemade karaoke, also has powerful applications across TV, film and the music industry.

Timestamps:

• 08:30 – Intro to AudioShake
• 15:45 – How the tech alters the content creation process
• 19:00 – Speed of processing and real-time
• 23:00 – The reaction of the sports industry
• 30:00 – Karaoke and the company's founding
• 33:30 – The business model  
• 39:30 – Real-time audio splitting 
• 45:30 – Training the AI models