From the Big Sort to Big Data: Tech and talent evolution drive managed review revolution

Tuesday, September 22, 2015 - 14:41

Shelley Brown, an attorney and the Director of Managed Review Services at RVM Enterprises, Inc., discusses the evolution of the business of eDiscovery. Controlling costs, proposed changes to the FRCP, big data and Watson all are influencing the industry. Ms. Brown gives us her take as someone who’s been on the cutting-edge of change in her 12 years in the field. She can be reached at

MCC: You began your career as a practicing lawyer. What led you down the eDiscovery path and, eventually, to RVM?

Brown: When I graduated from law school in 2003, I had never heard the word “eDiscovery.” I certainly hadn’t had any courses in it. I started out at a securities class action firm and was asked to write a brief on the sorts of metadata that need to be produced to a regulator. I had no clue what metadata was, let alone the implications for production, so I had to learn it, and fast.  I called up the regulator and asked.  That answer, and some other research, became my brief. That brief somehow became popular and people started asking me about eDiscovery. I kept learning about eDiscovery, by research and by doing, and I've been at it now for 12 years.

MCC: A lot has happened in those 12 years.

Brown: It has. When I started in 2003, you'd have a room full of people and boxes stacked from floor to ceiling. It was what I call “The Big Sort.”  We were creating huge stacks of documents that somehow were associated with one another. The process was entirely paper. Even where there was electronically stored information, we would print it out and sort it. I’ve had my fair share of paper cuts. Now the process is entirely different.

MCC: What impact has big data and big data analytics had on eDiscovery and managed review?

Brown: I often hear quoted the result of a UC San Diego study that suggests that 99.8 percent of all of the data that exists today was created in the last two years. This data deluge is reflected in the increase in volume of data subject to discovery, which typically accounts for 60-70 percent of the cost of litigation. There's always been a need to keep those costs down, but with big data that need has mushroomed. You have to employ different technologies and sophisticated workflows just to get through the data in the time allotted for discovery. In most cases, linear review just won’t work anymore. The focus has shifted away from a hunt-and-peck type of review to using technology to help you get to the most important data first, before the discovery clock runs out.

MCC: Corporate counsel have a strong interest in controlling/reducing costs, and in litigation that means discovery and eDiscovery. What can you do to help companies keep costs down?

Brown: We have a great ability to leverage technology and people at RVM. We have eDiscovery technologists and analytics experts in-house, and the first goal is to cull our clients' data down to exactly what is necessary. We try to get rid of as much noise as possible, identifying things that are never going to be relevant to litigation or discovery, like your Chinese food menu, and getting those out of the way to make it easier to find the smoking guns. 

Then the human element comes in. Our contract attorneys are real attorneys. They are graduates of ABA-accredited law schools, or an international equivalent, and are barred and in good standing in at least one U.S. jurisdiction. And those are just the basic requirements. We also go to great lengths to find specialists in the underlying subject matter, or people who at least have some experience with those types of reviews. The members of our team are as proficient as, and in some cases even more experienced than, junior associates in a law firm. The difference in cost is significant. In places like New York, for instance, it isn’t uncommon for a junior to mid-level associate to be billed at around $400 an hour. A contract lawyer may cost around $40 an hour, in certain circumstances, for the same caliber work. That's how we use human capital to help our clients save money, but there are often other advantages. Lawyers doing contract work often are technologically savvy because they work with eDiscovery tools day in and day out. We don't often find the same level of familiarity with eDiscovery concepts and tools with younger associates. Therefore, our clients are not only getting a lower-cost resource, but a very capable resource for about 10 percent of the cost.

MCC: Given what's been happening the last few years in legal education, particularly graduates’ inability to find legal jobs, has your talent pool changed?

Brown: It really has. The pool has grown because the market is saturated with attorneys who need jobs. Lots of people end up leaving the field altogether, but many who stay find document review to be a good way to generate extra income or to make a career. Many members of our team have private practices and use document review to supplement that income. It runs the gamut. Some work on one project for a year. Others are with us for years and rely on document review for their entire income stream. Others are recent graduates who have not yet been able to find steady legal employment. A lot of our millennials just like the flexibility. They work 90-hour weeks for four months, then take two months off. Rinse and repeat.

MCC: Is part of your role to be the gatekeeper for the human side?

Brown: We have a team of review managers on staff. Together, we are the gatekeepers. My goal is to make sure that we have defensible and repeatable processes in place for each matter and that the review managers implement them consistently. They understand the needs of the people on our team, as well as considerations on the law firm side. They’re well suited to liaison between our clients, whether corporate clients or outside counsel, and the review team.

MCC: How does quality control work given the sheer volume of what goes on in some of these cases?

Brown: We have a standard workflow based on proved methodologies for quality control that we implement for each review. Though the standard process is static, the more granular aspects are quite customizable. Depending on the underlying substance of the review and whether advanced analytics are being used, those processes will change somewhat to suit the clients' needs.

MCC: You said your contract attorneys generally have some big firm experience. What other qualities do you look for?

Brown: Dependability is certainly one, and a willingness to understand and embrace technology. One might think that your ideal contract attorney is a retired partner from a white-shoe law firm, but that tends not to be the case. Our best contract attorneys are those who love and understand technology, have a broad understanding of the law, and are familiar with the subject matter.

MCC: The technology in this area moves very quickly, especially predictive coding. I know you have a lot of tech people on staff, but are you developing your own technologies? Or are you using others' technologies and adapting them?

Brown: It’s a mix. We strive for the gold standard in every aspect of our service offerings. RVM uses a portfolio of proprietary products combined with best-of-breed applications. We leverage several analytics tools, and we've built many of our own advanced analytics workflows to operate in conjunction with these tools. We employ a mix of all of the above to come up with the best solution for each client.

MCC: From the client's view, cost is important, but they also want to get it right. Do you spend a lot of time educating clients and potential clients?

Brown: My job includes doing a lot of CLE presentations on trends in technology and the law. For example, last year there was a lot about the proposed changes to the Federal Rules of Civil Procedure. It will be very exciting to see where those go, especially the rules that pertain specifically to electronic discovery.

MCC: It looks like we're locked into December 1.

Brown: It does. Some of the debates are still quite interesting, specifically around sanctions for improper disposal of electronically stored information. I'm talking about Rule 37E, which I think is the sexiest of the proposed amendments. I'm excited to see where things shake out. Even though technology is all around us, and for me it's hard to think of the world without things like electronically stored information, our laws are still trying to catch up. The very first rule to contemplate eDiscovery was Rule 26. We're way behind the curve. Electronically stored information has been with us for so long: I remember sending my first email in 1994, yet we only began to really consider the evidentiary impact of that email in terms of law and procedure in 2006.

MCC: It's not really just the volume of information, which is mind-boggling, but the nature of that information. It’s incredibly complex.

Brown: We've had a data explosion, but we've also had a data source explosion. We used to talk about getting emails off one server. Now you have corporations with several legacy systems that don't talk to each other, and they all contain data that needs to be collected and reviewed. We also have to consider structured data, which was not traditionally part of managed review. The landscape of managed review is changing.

MCC: Are machines taking over managed review?

Brown: I worked on one of the very first predictive coding reviews several years ago. I had a large team of contract attorneys. It was not at all unusual in 2006 and 2007 to see managed reviews with hundreds of contract attorneys. When we started doing predictive coding, and rumors about its use started filtering down to the contract attorneys, there was a huge scare. Everybody thought, "There's no longer going to be contract attorneys. Everything is going to be done by computer."

Now it’s 2015 and predictive coding is no longer a black box. The courts say it’s okay to use, and we still have contract attorneys and managed review. It’s true we no longer have hundreds of attorneys sitting in a room clicking away very often. The sheer numbers of contract attorneys have been replaced – not just by technology but by a more informed reviewer – more informed about the subject matter specifically and the use of technology generally.

MCC: You've been steeped in eDiscovery a long time as a practitioner, a student and a teacher. Where are things going in the next 12 to 24 months? What's the next big thing?

Brown: I think IBM Watson could be the predictive coding of the future. They’re using cognitive computing to do amazing things in medicine, and some students in Toronto have created an App called ROSS that is built on Watson, but has been taught law. It’s important to point out, however, that systems like ROSS, if put into use, will still require human intervention, the same way that we require human intervention when we have a true predictive coding workflow. We can use machine learning to teach the analytical tools how to predict how documents ought to be tagged, but what we can't teach are the nuances involved with things like identifying work product privilege. Those are questions of law that require massaging and analytical interpretation that our machines, as far I've seen, are not quite prepared to do.

MCC: Machines can handle volume and more routine analysis, but not the higher-end work?

Brown: You want your lawyers occupied with the higher-minded legal analysis, and you can use Watson for the day-to-day analysis. Watson is not going to get tired; it won't need to take breaks. Usually with our contract attorneys we try to keep our projects to 10 hours or fewer because accuracy plummets after the 10-hour mark. Watson and tools like it may change the game.

MCC: I know you’ve thought a lot about the ethical implications of all this. Talk about that for a moment.

Brown:  When we're talking about Watson, or any sort of machine learning, it's important to remember a couple of things in terms of ethics. As attorneys, our Duty of Competence suggests that we have to acquire the knowledge of how these technologies work, and how to make managed review work, or we have to hire someone who has that knowledge. Technologies change, but our ethical obligations are constant.