# Music AI (Moises)

**Type:** venture
**Status:** Draft
**Confidence:** Medium
**Focus:** AI music tools, audio source separation, voice cloning, music generation, creative technology
**Stage:** Series A ($40M, January 2025; Connect Ventures / CAA+NEA, Kickstart, Samsung Next, Pelion)
**Location:** Salt Lake City, UT (founders from Brazil; US HQ in Utah)
**Updated:** 2026-06-19
**Needs-reviewed:** 2026-06-19
**Hero:** https://picsum.photos/seed/music-ai-studio/1600/1100
**Pull:** *50 million users, an Apple App of the Year, and a bet that AI should pay musicians — not replace them.*
**Relates:** cites [Official Website: Music AI](music-ai-official-website.md)

## Summary

Music AI (trading as Moises) builds AI-powered tools that empower musicians, producers, and creators — starting with source separation (isolating vocals, instruments, and stems from mixed audio) and expanding into licensed voice cloning, music generation, and collaboration infrastructure. The flagship Moises app has 50 million registered users and won Apple's iPad App of the Year award in 2024, reflecting genuine product-market fit among practicing musicians.

The more interesting long-term question is the ethical framing: Music AI is building infrastructure for artists to monetize their voice and musical style in the AI era, rather than having it scraped and replicated without consent. Artists have invested in the company alongside institutional investors, which creates alignment that pure-tech music-AI players lack.

## Impact

Source separation put professional-grade audio tools in the hands of 50 million people: remixing, learning songs by ear, building backing tracks, and composing from existing recordings became accessible to anyone with a smartphone. That's real creative democratization. The harder, bigger bet is whether licensed voice cloning and music generation can establish an industry-wide model where AI and artists coexist rather than AI displacing them. If Music AI's royalty and consent infrastructure becomes the standard, the counterfactual impact is significant; if source separation stays the core product, it's a useful tool with strong but bounded impact.

## What They Are Building

The core Moises app: source separation, stem extraction, chord and beat detection, AI backing tracks, and pitch/tempo manipulation. On top of that: Music AI is building a platform for licensed AI voice cloning (artists opt in, set terms, receive royalties when their voice is used in AI-generated content) and music generation tools aimed at producers and labels. The platform play is to become the infrastructure layer between AI music companies and the rights-owning music industry.

## What They Need Now

ML engineers with audio signal processing and generative model backgrounds; product engineers for the consumer app and platform; legal and business-development people who understand music rights and licensing; creative technologists and artists-in-residence; and data scientists working on audio models. The Brazilian founding team culture plus Utah HQ creates an international atmosphere unusual in the SLC tech scene.

## Who Could Help

Music-industry executives and label relationships who can validate and scale the licensing model; AI and signal-processing research partnerships; music-rights and copyright attorneys; streaming platform contacts (Spotify, Apple Music) who shape what tools are available to creators at scale.

## Utah Context

Music AI is one of a small number of Utah companies with a genuinely global consumer reach. It sits alongside creative-technology companies in Utah's emerging media and AI ecosystem. The Series A's Utah Kickstart participation signals local investor confidence; the CAA+NEA (Connect Ventures) co-lead connects it to Hollywood's talent ecosystem. Notable: artist investors including Steve Aoki, 3LAU, Freddy Wexler, and Alexander23 suggest the company has real credibility among working musicians.

## Evidence

- [Official Website: Music AI](music-ai-official-website.md)

## Open Questions

- 50M registered users ≠ 50M paying users. What is the paid-conversion rate and ARPU?
- How does the licensed voice cloning model work legally and technically? What prevents artists from being cloned without consent even within the platform?
- The source separation market is increasingly commoditized by open-source models. What is the durable moat?
- Imagery: hero is a picsum placeholder pending a license-clean press photo.