used active learning in Relativity, allowing a four-person review team to produce 235,000 documents in four weeks

A Canadian law firm had little time to spare on a case for a construction company in the resource and energy industry; they had four weeks to review 730,321 documents. And the firm had four reviewers to get through the large data set.

Pressed for time and hungry for ideas on how to proceed, the firm partnered with H&A eDiscovery to create an action plan.

Getting Active with Active Learning

Director of Client Services at H&A eDiscovery Lynne Brigant proposed using active learning in Relativity due to its speed and reliability.

"With active learning, case teams can quickly access vital case documents and make educated decisions,” Lynne said. “I advise our clients to use active learning for almost every case."

Lynne first reviewed the documents and excluded non-inclusive emails and non-searchable documents containing either insufficient texts or images, such as site photographs and computer-aided designs.

After eliminating 86,586 documents, the firm's attorneys ran the remaining documents through active learning. Active learning digested the reviewers’ coding decisions for relevance to echo these decisions on the documents in the review queue.

Small Team, Big Review? No Problem

Once the firm was confident in active learning’s coding decisions, the four reviewers looked through the documents pertinent to the matter.

With the team powering through the review, the firm produced 235,000 documents in a mere four weeks.

"When you look at how this small team reviewed hundreds of thousands of documents in that short timespan—it’s exceptional,” Lynne said. “Analytics is transforming e-discovery. It is an exciting time to be a part of the industry."

“When you look at how this small team reviewed hundreds of thousands of documents in that short timespan—it’s exceptional. Analytics is transforming e-discovery. It is an exciting time to be a part of the industry.”