How did they do it?
- Early Case Assessment, Analytics, and Assisted Review helped reduce the data set
- Saved their client nearly 8,000 review hours and $350,000
Preparing for Review
After a large company was confronted with allegations of violating the Employee Retirement Income Security Act, they needed to take a deeper dive into the relevant information. They turned to Legility to reduce the documents for review so they could assess if there was any merit to claims that they failed to disclose investment options’ expenses and risks.
Legility’s forensics team got to work and conducted a series of collections, which yielded 1.9 million documents spanning 17 custodians.
To cull the documents for review, Legility pushed the data into an Early Case Assessment workspace in Relativity and ran a search based on names and terms their client provided. Then they took an interactive approach. They ran multiple search terms reports and tracked the number of hits certain terms accrued. This allowed Legility to identify which terms appeared more frequently among the document set. Using this insight, they refined their original terms list to yield a more targeted list of documents, which reduced the document set by nearly 77 percent.
Most of the remaining 440,000 documents were emails. Legility ran email threading to assemble email content in an easy-to-review format and eliminate duplicative content. Threading found that 135,000 emails had no unique content and could thus be removed.
Putting Assisted Review to Work
Then, to further accelerate review, Legility approached their client about using Assisted Review to help automatically classify documents as responsive and non-responsive.
“We knew this case would be ideal to use Assisted Review with threading based on the sheer volume of documents,” said Natalie Franke, senior project manager at Legility.
Two subject matter experts sifted through the documents that were not eligible for Assisted Review, such as those containing insufficient text. Then, the experts coded a sample set of eligible documents and conducted a series of training rounds to ensure Assisted Review was accurately coding documents based on responsive criteria. Then, Legility finished running the full document set through Assisted Review. The results yielded 34,000 documents coded as responsive and 257,000 as non-responsive, with nearly 15,000 documents uncategorized. Legility bundled the responsive and uncategorized documents—just 49,000—for the client’s review team.
Using Assisted Review and email threading, Legility truncated the documents for review by more than 99 percent, saving their client 7,739 review hours and $348,261 in costs.
“This was a big success for us when you look at how we substantially reduced the data universe. Our client was excited about how we leveraged technology to get the job done. Using Assisted Review with email threading made that possible.”
NATALIE FRANKE, Senior Project Manager