Dr Alec Wright

What problem are you trying to solve?

Large‑scale audio ML models can unlock novel musician‑centred tools, from text‑to‑sound generation and timbre transfer to semantic audio understanding and intelligent audio effects control. Yet most music software companies are small teams without the research expertise, data pipelines, evaluation practices or compute to build and sustain these systems. 

This capability gap slows innovation and keeps useful ML features out of reach. I’m addressing this by developing sector‑specific audio foundation models and collaborating with industry partners to co‑design prototypes and co‑develop real‑world products for musicians.

What is your idea?

I’m developing sector‑specific audio foundation models and collaborating with music software companies to co‑design, prototype and pilot new products. Partners gain ready‑to‑integrate models, lightweight runtimes and example code, and we work together on evaluation and deployment.

The aim is to translate state‑of‑the‑art research into practical, musician‑centred tools without each company needing to build an ML stack from scratch.