OpenText’s Hal Marcus and Anthony Di Bello discuss how to make sure e-discovery is a part of an organization’s overall information governance and management program. Their remarks have been edited for length and style.
CCBJ: Over the past five years, OpenText has made a number of acquisitions, the most recent being Guidance Software last August. What has been the strategy behind these acquisitions?
Hal Marcus: OpenText is a leader in enterprise information management, and that means the entire lifecycle of information for corporate, government and other customers. A key need in the lifecycle is around insight, understanding what’s in the information. And that leads us to the legal process of discovery, to collection and analysis, and to the securing and protection of that information.
Anthony Di Bello: Customers are turning to OpenText to provide other required capabilities around sensitive data management, including machine learning and AI capabilities, to deliver deeper insight and meaning.
How do you differentiate yourself with your combined offerings?
Marcus: We provide unparalleled coverage of the EDRM (Electronic Discovery Reference Model) from the identification of critical content, through preservation and collection, and through all the processes, including production. We’re now both a best-of-breed and a combined, unified offering.
Di Bello: We’re now elevating this by leveraging our discovery capabilities into information governance and information lifecycle management as a whole. In this context, e-discovery is part of an organization’s overall information governance and management program.
What does a typical client look like, and what problems are they trying to solve?
Di Bello: For Guidance, that’s really a strong corporate and government customer base. Our focus is on visibility and collection, including searching and culling at the point of collection – things that are important to the corporate counsel controlling how much data they’re sending out to review.
Marcus: Historically, Recommind’s discovery opportunity would arise when a corporation or a law firm had a substantial amount of data they needed to get through, as quickly and as efficiently as possible, both to inform their case strategy and to keep costs low and get to the production stage.
Guidance’s reputation was built around its forensic collections technology. How have clients leveraged this technology?
Di Bello: The Guidance EnCase suite of products has become the most widely deployed search and collection technology within corporations. That’s because there’s a high bar of entry to agent-based collection solutions, given the complexity of maintaining support for different file and operating systems in a manner that ensures confidence of findings and the ability to gather up all relevant ESI.
Marcus: Seventy percent of the cost of litigation in the U.S. goes to discovery. And 70 percent of the cost of discovery goes to the review process. If you can cull your data intelligently right at the source, in a forensically sound way, then you can reduce those volumes. And if you can apply the best technology to get you through that review process efficiently, then you have a huge impact on your discovery cost and much faster access to the facts you need to inform your case.
Recommind built a reputation around its legal analytics and machine learning, including predictive coding. What’s your secret sauce, and how have your clients used this technology? Are they using it outside of traditional litigation contexts?
Marcus: The genesis of Recommind was not actually discovery or document review. The company was founded based on unstructured data analytics technology, a form of AI. It looks at the interrelationship of text within documents and across databases and performs a statistical analysis. From that context, it can derive meaning. And it does it in a language-agnostic way, without relying on taxonomies. The effect is that we can create concept groups, enable phrase analysis, and leverage metadata to search, filter and cull data.
Applying that to the discovery challenge is what led us to develop OpenText Axcelerate, which embeds these capabilities throughout the process. The infusion of AI creates huge efficiencies and brings the insights out earlier. That’s our secret sauce.
Di Bello: I’ve talked to a couple of customers, and one of the comments I’ve heard was that between the ability to cull at the point of collection with EnCase and the analytic insights provided by Recommind, they’re reducing their ESI volume by about 97 percent before sending it out.
Do you think that advanced analytics and machine learning are living up to their potential?
Marcus: In a word, no. And the reason is that the predominant approach to machine learning in the legal industry requires a detailed protocol and a lot of advanced setup to implement.
The most widely used platforms for technology-assisted review require a lot of heavy lifting up front, a lot of negotiation over protocols, a lot of agreement around statistics and structures. That hampers adoption significantly. We’ve done surveys where people are talking about the promise of predictive coding but indicating they don’t get to reap the benefit all that often.
That frustrates us because our technology can be applied flexibly at no additional cost, with no additional setup. Many clients of OpenText Discovery use machine learning on every matter, on virtually every data set.
Where do you think the e-discovery industry is going?
Di Bello: On one level, I see the industry evolving into one more focused on governance and management, in which the ability to find information related to litigation is important. The market for e-discovery has become part of a bigger legal risk management operation. Technology is moving in that direction, and that’s one reason OpenText is looking to ensure they have those governance capabilities around the data they’re helping their customers manage.
Marcus: We’re seeing a substantial breaking down of barriers around different use cases and the technologies that apply to them. This includes internal investigations, due diligence, data breach response, as well as litigation. Advanced analytics and machine learning bring similar value to all these use cases.
A related and continuing trend is greater corporate control over the data associated with discovery. There’s no need for corporations to send the data anywhere, just as there’s no need for them to have multiple copies of their most sensitive content floating around in different servers and among different service providers. They can keep this consolidated in a highly secure environment and control the access points – for outside counsel, for experts, for service providers – as needed.
Security is a broad and rapidly changing market. How will OpenText continue to address this area?
Di Bello: The acquisition of Guidance Software is really just the first entry into the security market, establishing visibility into endpoints, where sensitive information is stored. There are a couple other technologies we’re starting to explore from a leverage perspective so we can expand the ability to detect threats, respond to threats and ultimately eliminate threats to sensitive data.
If you look at the security market in general, folks are looking to leverage machine learning technologies given the skill gaps and skill shortages in the space. OpenText is focused on acquisitions to fill gaps, build businesses, pull technologies together.
And while we develop our overarching security strategy, OpenText’s advantage is the ability to bring security closer to the data. Most security companies are focused on putting walls and moats and barriers around an enterprise perimeter; however, remote workers, mobile devices and virtual networks have changed the perimeter. The perimeter is the data. And OpenText manages that new perimeter.