The music industry is facing a growing challenge: unpaid royalties for copyrights are expected to soon reach almost 8 billion dollars annually, compared to the current 2.5 billion dollars. In a groundbreaking initiative to combat this unfair distribution, the Eastern Swiss start-up HELGA.works is launching an Innosuisse project with the Chair of Data Science & Artificial Intelligence at the University of Liechtenstein.
The aim of the project is to help artists receive the royalties to which they are entitled. The main challenge is to identify the rightful recipients despite insufficient data quality, a problem that affects niche artists in particular, but not only. By using state-of-the-art machine learning models, such as graph neural networks, which can deal with incomplete and inaccurate data, the aim is to create a solution that is not only innovative, but also remains simple and interpretable. The aim is to develop a prediction service that estimates the risk and amount of unpaid royalties.
The potential product solution has already attracted significant interest from investors as well as the industry, including renowned music publishers and artists, and promises a competitive advantage for HELGA.works and its clients. Even rough predictions can be very helpful in guiding efforts to further analyze potentially unclaimed royalties. Based on a model prediction, an agent can decide whether to further investigate the causes and potential amounts of unpaid royalties of existing clients or to focus marketing efforts on potential clients who would benefit most from HELGA.works’ offerings.
In addition to creating economic value, the proposed solution also supports niche artists in particular in receiving their rightful remuneration and thus contributes to musical and cultural diversity. Together with the University of Liechtenstein as a research partner, a Swiss start-up is involved in the Innosuisse project as an implementation partner whose employees have knowledge and experience in the field of machine learning and data science.