EIDR Modernizes Deduplication Efforts

For more than 15 years, the Entertainment ID Registry (EIDR) has been using rules-based deduplication to ensure content uniqueness by automatically comparing new asset metadata against existing records.

And while it’s served its purpose, the technology required regular re-programming to keep it current.

In 2026, EIDR has updated its deduplication technology, adding artificial intelligence and machine learning that evolves and improves on its own. Manual review rates are already significantly lower, and Hollie Choi, managing director of EIDR. said that within days of implementing the technology, EIDR users reported fewer registrations requiring manual review.”

“There was an immediate benefit to the Registry,” she said. “As we go in and make decisions, it will learn and start making decisions for us. It was time to move to an AI, ML system to be future-ready. It’s a major leap forward for EIDR, adopting newer technology that allowed us to scale, and scale quickly.”

Choi likened the tech update to building a multi-lane highway with future traffic in mind, futureproofing for the content traffic EIDR expects in the years to come.

“Moving to a human-in-the-loop machine learning system gives us the best of both worlds: AI-driven de-duplication that carries the bulk of the load with human review for the edge cases,” said Richard Kroon, EIDR’s technical director.

The technology comes via AI-native master data management (MDM) solution provider Tamr. Tamr has processed billions of records for global customers, reflecting the scale at which organizations today are relying on AI solutions to power critical systems and new and existing initiatives. Tamr said that API web requests nearly tripled year-over-year, as more enterprise systems looked to Tamr for upgrades to their operational needs.

“We're not just taking a general-purpose large language model (LLM) AI and pointing it at our records,” Kroon said. “The Tamr ML system was built from the ground up to perform metadata record de-duplication, and it does that one thing very well.”

Tamr CEO Anthony Deighton added: “Across industries, organizations are moving beyond AI experimentation, but AI in production is a different game than AI in a pilot. To get the most value from AI systems, the data behind them has to be unified, accurate, and always up to date. Our record growth shows the power of our AI-native approach to data mastering — helping companies move beyond rigid, rules-based approaches to build trusted, connected foundations that keep up with the demands of their business.”