Keep up with all of the essential KM news with a FREE subscription to KMWorld magazine. Find out more and subscribe today!

The secret of the cloud: Remote collaboration, elasticity, and the e-discovery paradigm

Article Featured Image

Cloud computing has quietly emerged as the focal point of e-discovery. This is significant for two reasons. The first is the domain’s horizontal reach, which any organization may become embroiled in, especially since it now transcends litigation to also include data privacy regulations, internal investigations, and Freedom of Information Act requests. The second is that e-discovery encompasses so many facets of the data management ecosystem—advanced analytics, data governance, metadata management, data visualizations, data cataloging, and more—that it’s an undeniable reflector of where that ecosystem is today: in the cloud.

“A lot of data in any corporation, in any entity really, is in the cloud,” said AJ Shankar, CEO of Everlaw. And, if it is not already in the cloud, he noted, it is moving there. Cloud computing and architecture provide the following three advantages for the e-discovery market and, by extension, IT resources in general:

Remote collaboration enablement: The controlled, yet remote, access to the cloud supports the sort of collaboration that’s become central to distributed communication, particularly in the past several months. “If there’s a law firm you’re working with for outside counsel, there’s a whole gamut—partners, associates, IT folks, paralegals, litigation support—a whole ecosystem of people who need to be in this platform, especially across multiple organizations,” Shankar noted.

Elasticity: The cloud’s ability to provision compute power on demand for computationally intense workloads such as cognitive analytics (which can be scaled back to address cost concerns) is almost impossible to duplicate on-premise. “A SaaS provider by its very nature has accommodated its computing resources to be able to spike to quickly apply a machine learning model, and then pull back to simply review a document,” said David Carns, chief revenue officer of Casepoint.

Scalability: In addition to the elasticity benefit of the cloud that Carns referenced, there is horizontal scalability that supports such resource provisioning. This correlates to its cheap storage whereby organizations can inexpensively store “very large sets of data which impact 20 different business units within a company, maybe 50 or 60 different applications, and hundreds of thousands of users,” observed Michael Jack, vice president of global sales at Datadobi.

It is clear that e-discovery is a microcosm of data management’s macrocosm. Winning in the cloud with the former illustrates how to do so with the latter.

Advanced analytics

Although e-discovery analytics may not be the most sophisticated, it offers a core value proposition that most organizations can’t afford to forgo. Implemented in the cloud, e-discovery analytics enables business users to quickly find the most meaningful data to help them do their jobs. These platforms facilitate the Technology Assisted Review (TAR) process via supervised learning. In this process, end users train machine learning models by manually indicating the fraction of a corpus that is relevant to their case. This approach enables companies to reduce the time and cost of looking for discoverable documents by “30% to 50% because of the enrichment, the analytics, and the AI prioritization,” explained Kiwi Camara, CEO of DISCO. “On a million-dollar case, you might save $300,000.”

End users’ labeled examples form the basis of these models’ ability to find similar results at scale. They also prioritize results based on confidence scores to “identify the really relevant ones for your case,” Shankar observed. This capability is critical for litigation because “many legal cases settle before they go to trial,” Carns explained. “In fact, I would argue the majority of them do. The trick isn’t to be exhaustive on Day One,” he said. Instead, the idea is to gather the most salient pieces of information on Day One in order to make a decision about whether it would be beneficial or not to settle a case early. This ad hoc, self-service paradigm of training AI models to inform immediate action is an approach that is applicable to any business user who needs to make datadriven decisions.

KMWorld Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues