Every major technology has two phases. The piano started as a parlor toy and became Carnegie Hall. The internet started as AOL chat rooms and became the backbone of banking, healthcare, and government. The phase-two builders are the ones who got rich. AI music is at AOL right now. We build phase two.
In 1787, Mozart's Musical Dice Game algorithmically generated unique minuets by rolling dice. Generative music is 240 years old. We're not reinventing it — we're finally finishing it.
The Problem
Live bands play flawlessly together, but syncing with stage tech still requires endless, manual rehearsal.
Music lawyers understand copyright law, but they cannot decipher the complex AI claims made by tech corporations.
Labels want custom AI tools to monetize their catalogs, but uploading assets to big tech risks their data security.
We bridge the gap between cutting-edge AI research and practical musical application for creators, performers, and the industry.
Producers & Songwriters
We build custom studio tools and help navigate the messy world of AI monetization and watermarking, ensuring you can commercialize AI-assisted tracks on your terms.
Live Artists & DJs
Our focus is the stage. We build lightweight software for real-time AI generation and live, interactive audio experiences that run on a single laptop.
Labels & Legal Teams
We build proprietary AI generators trained strictly on your own catalog, while providing the technical and copyright advisory you need to protect your assets.
Our core capabilities
We sit down, listen to your specific bottlenecks, and figure out exactly what kind of solution you need — no jargon, no pitch deck.
We design a tailored plan scoped to what you actually need — lightweight where possible, robust where it matters.
We build it, integrate it, and train your team on it. You own everything — no dependency on us to keep the lights on.
Track record across infrastructure, legal compliance, and live performance technology.
End-to-End AI Generation & Training Pipeline
Built a custom text-to-music model alongside an LLM-verified data annotation platform — where an LLM validates human-labelled data before it enters the training pipeline.
Translating AI Research into Legal Defense
Acting as technical translators for PRS lawyers, decoding bleeding-edge AI research to prove the exact financial value of human music used in corporate training data.
Real-Time AI Gesture Control
A laptop-only framework that learns a performer's body movements in real-time, mapping gestures to live DAW controls. Published at the International Computer Music Conference 2025.
Everything starts with a conversation. We'll listen to your specific bottlenecks, identify what kind of solution fits, and outline what working together could look like — no commitment required.
Secure scheduling via Calendly
Two engineers who are also working musicians. That's not a coincidence — it's the whole point.
Ali Bahrani
Ali brings deep software engineering expertise to every engagement, with a focus on building robust, production-grade AI systems. As a vocalist and musical theatre performer, he understands what artists actually need — and what they stand to lose. That duality drives every technical decision he makes.
Mahya Khazaei
Mahya specialises in machine learning systems for audio and music, from custom model training pipelines to real-time inference at the edge. As a pianist and composer, she approaches every model with an ear for what makes music feel human — and the technical rigour to protect it.