In a case involving a group of major commercial airlines, the defendant’s eDiscovery solution used an add-on analytics feature. The Technology Assisted Review (TAR) Control Set estimated their production would contain 85% of all responsive documents with a 58% precision rate. With these estimates, the defense production was expected to contain at least one responsive document for every non-responsive document captured by TAR (1:1). The Defendant produced 3.5 million documents by the strict deadline provided by the courts.

Plaintiffs Use Casepoint Active Learning to Verify an Overproduction

Due to the volume of documents, the Plaintiffs decided to reanalyze the production in the middle of the case using Casepoint’s legal discovery platform with powerful built-in legal AI software. Taking a Validation Sample, the Plaintiffs used CaseAssist active learning and advanced analytics suite to audit the data. Next generation TAR 2.0 features discovered that the original TAR Control Set estimates provided by the Defendant differed from the Validation Sample produced using Casepoint. The results of the TAR 1.0 versus the TAR 2.0 analysis concluded the initial production actually contained over 97% of all responsive documents, but the precision estimate was significantly lower, only 16.7%. 

This changed the responsive document review ratio from 1:1 to 4:1. With 3.5 million documents in the production, the new data analysis gave the Plaintiffs reason to claim an overproduction and follow with a motion to extend. Despite the objections, the Defendants’ attorneys reviewed and confirmed an error in their control set estimates. Following the Defendants’ acknowledgment of the error, the Plaintiffs motion to extend was granted by the Court.

CaseAssist Active Learning: Accelerate Time to Insight

Casepoint’s artificial intelligence and advanced analytics, collectively Called CaseAssist is built-into the platform. CaseAssist helps our clients identify the most relevant documents and cull non-relevant data quickly.  Dynamic training sets and Active Learning leads to superior outcomes when compared to using multiple products and bolt-on analytics software. Adopting legal technology with built-in AI capabilities can streamline workflows and accelerate a user’s time to insight by as much as 83% with improved accuracy and less resources. The Plaintiffs, in this case, were able to perform their audit swiftly, verify the overproduction, and win their motion to extend.

New call-to-action