THE PREMISE
Music has geometry. We build inside it.
Pythagoras heard it in ratios. Kepler mapped it to planetary motion. Every composer since has been navigating the same underlying space; the relationships between notes, harmonies, rhythms, and arrangements. This determines whether music feels coherent or arbitrary, resolved or suspended, familiar or improvised.
The result is music that is not only listenable, but traceable. Every note connected to the logic that produced it, every result ready to be used as more than sound.
ORIGIN
Where it started, and why.
The Structured Harmony Engine began with a long-running fascination: music theory, fractals, experimental composition, procedural art, and algorithmic performance tools. That work sat alongside years inside the music industry, where we built metadata systems and recommendation infrastructure.
From that vantage point, two pressures became clear.
- AI content was piling up on the work of real artists, and the teams training those systems had no clean, structured, original ground truth to learn from.
- The models themselves were closed. What the field needed was transparency, an engine you could explain down to each decision.
The question that followed was direct: could a system understand music deeply enough to compose it, mix it coherently, and stay original without drawing on any specific artist's recordings? The answer drew on research into structured data, the practice of abstracting a domain into its component parts.
With backgrounds in both data and machine learning, we concluded that the strongest path was to own the engine outright, with full control over its mechanics, rather than inheriting someone else's work.
MARKET OPPORTUNITY
Three large markets. One structural gap.
The companies building AI and music technology, games, enterprise audio, and learning tools share one need: music that is structured, that scales, and that arrives with an account of how it was made. SHE was built for exactly that.
AI Training Data & Music Technology
$4B+
Demand for labeled, structured, original audio for model training and evaluation.
Game Audio
$3B+
Studios need non-repeating soundtracks that scale across sessions and expanding content.
Enterprise Background Audio
$2B+
Retail, telephony, hospitality, and branded spaces that need continuous, catalog-free music.
Each market is served today by licensed catalogs, loop libraries, or production on a fixed schedule, all of which buckle at the scale and specificity these products demand. Across AI, gaming, and enterprise audio, the immediately relevant spend tops $9 billion, with education adding more.
Market figures reflect current spend on music licensing, AI training data procurement, music data and tooling, and game audio production across their respective industries. Sources available on request.
STRUCTURAL ADVANTAGES
Three properties that reinforce each other over time.
Built by hand, original by design
People designed the musical knowledge, the catalogs, the styles, the logic, and the engine, from first principles. SHE learned from no one's music, so every piece is original, with full provenance on each one, and clear to use in a commercial product. It isn't a feature bolted on later; it is how the engine is built.
A system that sharpens with use
The chord patterns, rhythms, and styles deepen with every cycle. Each reviewed piece tightens quality and coverage. The library improves the engine, and the engine improves the library.
One engine, many markets
AI and music technology, gaming, enterprise audio, and education all run on one core, set to different styles. A gain in one market carries to the rest.