Your single source for new lessons on legal technology, e-discovery, and the people innovating behind the scenes.

Our Secret to Engineering for e-Discovery

Sam Bock

Our team, just like our software, is built to support the quickly evolving needs of e-discovery. Tightly integrated product, engineering, and marketing teams strive for alignment on industry demands and our roadmap—but to reach that goal, we need to stay as educated on this industry as the professionals in the field who are pushing themselves to manage every case as effectively and efficiently as possible.

So how does an e-discovery software company keep pace with the push and pull of case teams’ unique and changing needs? To find out, we turned to our own and sat down with Nick Robertson, our vice president of product and marketing, and Perry Marchant, vice president of engineering at kCura.

Sam: What makes e-discovery ripe for innovation?

Nick: First, it’s a newer industry—I’d argue e-discovery as we know it is less than 15 years old. We’ve come far, but this space is still maturing, and that leaves a lot of room for innovation. It’s also because we’re dealing with a lot of macro tech trends: we’re trying to gracefully handle billions of records, while applying data analytics and machine learning.

None of that is unique to e-discovery, but we need to meet these challenges in very specific ways. For example, in a lot of spaces, search means returning the most relevant hits, but not everything. In our world, searching for something means you really do need everything. In other industries, a “record” might be a row of data; for us, each one is an actual document. All those millions of entries in our databases are files full of text—so they take up a lot of space, and it’s a lot of unstructured information to explore.

Perry: To tie a few of those things together, the amounts of data are growing so quickly that some of the things Nick mentioned—data storage, record retrieval, analytics—force us to take a deeper look at the problem and the solutions. Even when we have a solution in mind, we have to be critical of it from the start. This means thinking through questions like how will it scale? How can we make it sophisticated enough to grow with the space?

Look at the auto industry. There are a lot more cars on the road today than there were 50 years ago. More cars means more potential for accidents, so plenty of safety improvements have come along to support that evolution in the last five decades. But we need to reinvent our software every few years to keep up with the needs of this market. It doesn’t take 50 years or even 5 years for our industry to scale to the point that we need to reevaluate our approach. It’s a very compressed cycle, and we constantly innovate because of that.

Which is the bigger driver for change: emerging technology or evolving e-discovery needs?

Perry: There are definitely examples to support both points. In some ways, the technology available is changing e-discovery. But sometimes it’s the other way around.

Nick: I like to ask our customers what kinds of problems they’ll need to solve in the next two years, and what software like Relativity will need to do to help them solve those problems. I often get the response that the technology is ahead of their process, and they want to take more advantage of what they have today. In that sense, I see the technology being a bigger driver for change.

At the same time, our customers are facing all kinds of challenges and they’re looking for new ways to handle them. The proof is in our online ideas portal.

Perry: I’d agree. If we’re talking about where the tech is being pushed by the needs of the space, I think Relativity Data Grid in the 9 series is a great example. We’re pushed to build it because the data is just so big. But I also look at analytics, where it’s largely the technologists saying “Hey, this offers a better way to perform e-discovery—it’s faster, it’s more accurate, and it’s just helpful.”

How has the e-discovery industry surprised you?

Perry: It’s not an e-discovery specific thing, but the size of the data has shown even crazier growth than I expected. At the same time, the relatively slow adoption of analytics has really surprised me. We have plenty of customers licensing it, but it’s not being applied to as many cases as I expected by now. When I joined kCura in 2010, I pretty much thought manual linear review would be done in 12-18 months. Who’s going to do it that way when you have the analytics option? Pickup of those tools has been much slower than I would've guessed, from my view as a geek in this space.

Nick: What surprises me is how quickly the industry has matured. When kCura came into this space, we were basically a garage effort—a few of us in a small room, totally bootstrapped. Now to compete you have to be a legitimate software company from the outset. When I started in e-discovery over 12 years ago, we’d find CDs in boxes of documents, and try to convince case teams to discover the files on them instead of scanning the CD and Bates stamping it. We’re a long way from those days.

Perry: For you it’s gotta be like going from the typewriter to the word processor. You were doing what?!

Nick: I never used a typewriter, but I’ll take your word for it…

How does kCura keep up?

Nick: Two ways: 1) focusing on our customers—getting into the field and putting systems in place to capture, evaluate, and act on their feedback; 2) constantly re-evaluating how we operate and working on the agile process. It’s constant change, period. It’s how it has to be.

A few more specifics on number 1: We have a huge focus on events and staying engaged with the community. We send large contingencies to industry conferences, host our own conference in the fall and a roadshow every spring, and conduct perception surveys to ask customers for feedback and gauge their feelings on the industry. We perform a detailed analysis of every deal, won or lost, so we can learn from it. And, as mentioned above, we have an online portal where people submit ideas for the software and interact with product managers.

We do hundreds of roadmap presentations annually to share our plans for future development of Relativity, and we have team members dedicated to analyst relations and customer wins so we can get insight into what’s working and what isn’t in the field. Also, Andrew is always reporting on what’s going on, getting in front of customers, and driving us to do better.

Perry: On the engineering side, we update Relativity monthly to keep up with everything we’re learning about the industry. Our developers are on dedicated teams for each product across the stack, and they work closely with product managers and marketing to execute on a timely roadmap for our users. Major releases of the software come out twice a year, and we share test releases during the construction process to get internal folks using the software and providing feedback. We’ve also added a lot of automated testing in the last year or two.

More broadly, we frequently engage external experts to help us understand new agile best practices and new technologies, like data storage options, hitting the market. We keep tabs on the evolution of the technologies we build into our software as well as emerging technologies.

It’s easy to get stale if we’re not constantly reevaluating our approach; we always need to be aware what supporting technology is best-in-class today, and not assume one decision about storage options or UI design makes sense forever.

Nick: Ultimately, the magic in this is when you deliver something your users didn’t think to ask for but makes their lives a lot easier. And then they start using Relativity in ways you never imagined. That’s a real thrill for us.


Nick and Perry will be hosting this year’s Innovation Awards at Relativity Fest, recognizing the community’s best custom solutions built on Relativity. Learn more about the finalists and vote for your favorite here.

Nick Robertson is vice president of product and marketing, where he guides the product roadmap and oversees marketing strategy. Prior to joining kCura in 2007, Nick earned experience in electronic evidence and e-discovery at a computer forensics firm and a leading provider of e-discovery services.

Perry Marchant is vice president of engineering at kCura and leads the evolving design and functionality of Relativity. Prior to joining kCura in 2010 as a senior software engineer, he developed security solutions for Symantec and McAfee and served as a consultant for Fortune 500 companies.


Sam Bock is a member of the marketing team at Relativity, and serves as editor of The Relativity Blog.