In 2015–2017, the generative music market barely existed. Our idea was born from a hypothesis that electronic music could be algorithmically generated based on context (e.g., for running or focus) rather than just recorded. We launched Mubert.com.

Design and launch an AI-generated music product from scratch: from developing the music algorithm logic to raising initial investments and building a professional team. Create an endless stream of copyright-free music.

Algorithm Design: Together with the CTO, developed the music generation logic that linked mathematical algorithms with music theory. Personally selected and classified thousands of audio loops for the initial database.

Fundraising & Pitching: Presented the project to investors and funds. The project gained recognition at major festivals (e.g., Alfa Future People).

Team Building: Recruited key specialists to the project, including the future CTO, music director, and database architect.

Product Analytics: Set up analytics systems to track early user behavior and validate Product/Market Fit hypotheses.

Monetization Strategy: In the early stages, laid the foundation for a B2B model (Generative Audio API for business), which later became the company's main revenue driver.