Loading organizations...
§ Private Profile · San Francisco, CA, USA
AI audio technology separates recorded audio into stems for the music industry, enabling remixing, spatial audio, and sync licensing.
Based in San Francisco, AudioShake develops artificial intelligence audio technology that separates recorded tracks into individual components, known as stems, to facilitate remixing, remastering, and sync licensing. The company provides B2B enterprise licensing and API access to major record labels, music publishers, and film studios, alongside a subscription platform for independent artists. The enterprise has raised $2.7 million in seed funding to date, securing its most recent capital injection in May 2023 to expand its engineering capabilities. AudioShake serves prominent entertainment industry customers including Sony Music, Warner Music Group, and Disney, and its technology was notably utilized to isolate vocals for a spatial audio mix of a classic Nina Simone album. Its financial backers include Indicator Ventures, Precursor Ventures, and Metallica's Black Squirrel Partners. The organization was founded in 2020 by Jessica Powell and Luke Miner.
Audioshake has raised $19.0M across 3 funding rounds.
Audioshake has raised $19.0M in total across 3 funding rounds.
Audioshake has raised $19.0M in total across 3 funding rounds.
Audioshake's investors include Cherry Ventures, Cross Culture Ventures, Indicator Ventures, Shine Capital, Thomson Reuters Ventures, Bradley Horowitz, Charlie Songhurst, Roneil Rumburg, Troy Carter, Felicis Ventures, Golden Ventures, Graph Ventures.
Audioshake has raised $19.0M across 3 funding rounds. Most recently, it raised $14.0M Series A in September 2025.
# AudioShake: AI-Powered Audio Separation Technology
AudioShake is an AI-powered music and audio technology company that uses advanced machine learning to separate any audio recording into its individual components, or stems.[2] Founded in 2020 and based in San Francisco, the company transforms audio from a static, "read-only" asset into actionable, editable data by isolating dialogue, music, instruments, sound effects, and individual speakers from complex audio files—even decades-old recordings that were never designed to be separated.[5]
The company serves a diverse range of industries including music, film and television production, sports broadcasting, podcasts, gaming, and AI companies.[2] AudioShake's business model combines annual access fees with usage-based pricing, and it has achieved significant traction with enterprise customers including Universal Music, Disney Music Group, Warner Music Group, Warner Bros Discovery, BET, NFL Films, and several major technology companies.[5] The company has processed over 100 million minutes of audio and achieved nearly 400 percent year-over-year revenue growth, with 40+ enterprise contracts driving its expansion.[5]
AudioShake emerged from a fundamental insight: most of the world's audio was effectively "read-only," accessible only for listening rather than editing or manipulation.[5] The company began as a music industry application focused on using AI to separate audio tracks—isolating vocals, drums, guitars, and other instruments—to enable remixes, karaoke, localization, and sync licensing.[1] This initial focus on music production revealed a much broader opportunity: the same technology could isolate dialogue from complex soundtracks, remove copyrighted music to avoid licensing issues, and separate overlapping speakers in podcasts and broadcasts.
The company's early success attracted significant investment and partnerships. In 2024, AudioShake won the top prize at the AWS re:Invent Unicorn Tank competition and became a key infrastructure partner with major artists like Green Day and companies like Disney Music Group.[2] This recognition validated the company's approach and accelerated its expansion beyond music into film, sports, and AI applications.
Broadcast-Quality Results on Any Audio: Unlike previous audio separation technologies, AudioShake works on any existing recording—including decades-old content and physical-world audio never designed to be separated—and delivers professional-grade quality through simple APIs.[5] This capability has enabled use cases like resurrecting classic recordings from artists like Nina Simone and Whitney Houston in immersive formats.
Comprehensive Audio Intelligence: The platform goes beyond simple stem separation. It can identify and isolate individual speakers even when they're talking simultaneously, transcribe 40+ spoken languages with sub-1-second latency, and align lyrics with 97% accuracy.[1] This multi-modal capability makes it valuable across diverse workflows—from dubbing and accessibility to voice AI and content localization.
Developer-Friendly Infrastructure: AudioShake offers a no-code solution that accepts live feeds in multiple formats (RTMP, HLS, SRT, MP4) and provides sub-3-second latency for both audio and caption creation.[1] The platform's simplicity and speed make it accessible to enterprises without specialized audio engineering expertise.
Proprietary, Patented Technology: The company has developed state-of-the-art lyric transcription and sound separation models built on self-supervised machine learning that adapts to different accents, dialects, and acoustic environments.[1][4] This technical moat, combined with millions of minutes of processed audio, creates a competitive advantage in model quality and reliability.
AudioShake is riding several converging trends that amplify its impact. First, the explosion of multimodal AI has created demand for high-quality audio understanding and manipulation—a capability previously locked away in mixed recordings. Second, the shift toward personalized, localized, and immersive content (Dolby Atmos, spatial audio, international dubbing) requires the ability to isolate and recombine audio elements with precision. Third, the rise of voice AI and synthetic media has created urgent needs for clean speech separation and speaker isolation.
The timing is particularly significant because AudioShake addresses a fundamental limitation in how digital media is created and consumed. As one investor noted, the company is "building the foundational layer that makes audio as flexible as text or images."[3] This positions AudioShake as infrastructure for the next generation of content creation, where audio becomes programmable and interactive rather than static.
The company's influence extends across the entire media and entertainment ecosystem. Film studios use it for post-production and foreign-language dubbing. Music labels leverage it for catalog restoration and immersive remixing. Sports organizations use it to remove unlicensed music and avoid copyright fines. This breadth of application suggests AudioShake is becoming a critical utility layer in professional audio workflows.
AudioShake is well-positioned to become the dominant platform for audio separation as AI-driven content creation accelerates. The company's combination of technical superiority, enterprise traction, and expanding use cases creates a powerful flywheel: more customers generate more training data, which improves model quality, which attracts more customers.
The next frontier for AudioShake likely involves deeper integration into creative workflows—moving beyond API-based separation toward end-to-end content creation platforms. As interactive audio, spatial formats, and personalized content become standard, the ability to programmatically manipulate audio will shift from a specialized capability to a fundamental requirement. AudioShake's early leadership in this space, combined with its growing enterprise relationships, positions it to shape how the industry thinks about audio in the AI era.