Music streaming services receive a huge number of tracks, with ~120,000 songs uploaded to Digital Service Providers (DSPs) such as Spotify each day (up from 93,000/day last year and 45,000/day in 2018). This incurs a huge administrative overhead for Artists and Repertoire (A&R) administrators in record labels, who must collate credits data detailing the songwriters and musicians who contributed to the recordings, which should be associated digitally with the tracks to enable allocation of royalties. Existing processes are non-standardised, and based on a mix of various unstructured data sources such as email, WhatsApp messages, Word documents and spreadsheets. This makes the process time-consuming and error-prone. Consequently, credits data is often incomplete or incorrect, especially for small independent record labels. The minimum metadata requirement for a song to be played via an online streaming service is a title and a unique identifier for the track (an ISRC, International Standard Recording Code); in practice, this is often all that is supplied. As such, the rightful recipients of royalties often cannot be identified. These monies are referred to as the "black box" of unallocated royalties, estimated by a 2019 US Government Accountability Office report to account for 20% of all online streaming royalties. They get divided up among record labels and publishers based on market share, and as independent artists and small independent record labels typically have close to zero reported market share, they receive little of these "black-box" payments. In 2021, the DCMS Select Committee ordered an inquiry into the economics of music streaming, calling for a "complete reset" so that songwriters, composers and artists, who currently receive "pitiful returns", are fairly rewarded. This project creates automation solutions employing AI to ensure complete, accurate metadata is associated with recordings so that royalty payments reach the appropriate rights-holders, fairly benefiting all stakeholders.