Media Company Saves US$1.5M and Meets Regulatory Deadline Using CAL and Agentic AI to Classify Large Dataset

Collect and review over 12 terabytes of data for a multi-party commercial transaction involving hundreds of custodians in the US and Canada.
Require extensive document coding to indicate category responsiveness across the entire production.
Face strict regulatory pressure to complete document production within 60 days.

Use Continuous Active Learning (CAL) to score and prioritize review of the highest-ranking documents.
Conduct privilege screening and specification coding on documents ranked highly in CAL.
Build models in Epiq AI Discovery Assistant™ to accelerate scoring and classification.
Apply review protocols to generate prompts that capture each specification and train the models.
Measure recall and precision for each model using existing coding.
Perform a linear review on less than two percent of documents not classified for any specification.
Pass documents classified by Epiq AI Discovery Assistant™ through privilege screening and CAL elusion testing to validate and complete coding.

Dedicated team experienced in designing technology-enabled processes for complex transactions.
First-of-its-kind AI platform accelerates review speed and improves accuracy.

Meet the 60-day regulatory deadline and requirement.

Saved US$1.5M by combining CAL and Epiq AI Discovery Assistant™, avoiding the cost of a full linear review.

Increased accuracy on complex issues missed by human reviewers.