The North American rights management org supports 150,000+ songwriters, composers, music publishers, and visual artists, licensing the performance and reproduction of music and the reproduction of audio-visual art.
Business objective & project requirements
- Improve data quality and data integration across 40 million master records siloed in the customer’s older, overly complex legacy system. The legacy system’s performance proved inadequate to the task of managing increasing volumes of new membership registrations over time, especially given the entertainment industry’s expanding deployment of new streaming and internet channels for music/video consumption.
- Dramatically reduce the time required to calculate and deliver performance royalties to members. This required real-time data tracking across internet entertainment channels (i.e., Spotify, Amazon Music, YouTube, etc), plus the ability to more accurately and quickly provide data quality checks across all data entities and attributes.
- Map usage data (song play transactions) to non-usage data (core master data, such as composers, titles, publishers, etc,).
Initial data challenges
- Consolidating and interrelating three business domains with 25 million musical works, eight million AV works, and eight million interested parties (songwriters, composers, music publishers, visual artists and licensees).
- Consolidating 38 tables, 2,563 attributes, and three levels of hierarchies, including a) people (creators, publishers, licensees,) b) things/assets (titles, running times, royalties splits), and c) royalties payment transaction data.
- Providing the backend power to manage and implement 1 billion+ rows/records created through consolidation and entities/attributes realignments.
Solution & Phases
Ataccama deployed an MDM (Master Data Management) solution to hierarchically link and consolidate all relevant membership master data, including data representing musical works, audio-visual works, and their respective owners across the organization’s creative community. Going forward, the solution will provide a backend engine powering real time data quality checks.
Phases:
- Create a canonical model for all attributes across the three domains to standardize all legacy system data formats.
- Build data flows into the MDM hub from 5 different input systems via batch and streaming interfaces, where some data is interchanged with the MS Azure messaging platform via JMS APIs.
- Expose golden records to consuming systems via web services, including both SOAP and RESTful formats.
- Create a master layer supporting upwards of 600 million records, including instance data from all sources, totalling 1.5 billions records in the MDM system.
Results
- Faster calculation enables royalties paid within days of plays, not monthly to quarterly batches
- A rule-driven, data quality solution that streamlines ongoing data quality activities
- Eliminated legacy system performance issues
- Powering quick searches for the client’s proprietary membership portal
