Scott Reents, Lead Attorney for Data Analytics and E-Discovery at Cravath, Swaine & Moore.
Scott Reents doesn't have what you might consider a typical lawyer background. He dabbled in computer programming while interning at a bank during college, and after earning an economics degree, veered into the tech industry during the boom of the 1990s. He later founded a nonprofit that developed voter guide software for media organizations, before ultimately turning to law. I've always been drawn to areas where the traditional way of doing things is breaking down, Reents said.
For someone attracted to disruption, Reents has landed at an unlikely law firm: the storied white-shoe firm of Cravath, Swaine & Moore. But as the firm's lead attorney for data analytics and e-discovery, Reents sees himself on the forefront of the legal profession's most important and exciting changes. He spoke recently with The Recorder about the skills required of e-discovery professionals today, how big data requires lawyers to think differently, and what legal tech vendors need to do better to serve their law firm clients. This interview has been edited for clarity and space.
Ben Hancock: E-discovery is changing. What skills do you need now to succeed in this field that maybe even five years ago you didn't need?
Scott Reents: Probably what most people would say is that the technology is increasingly sophisticated and increasingly an important part of the job of e-discovery, which is true. But I think a level deeper is a move from thinking about e-discovery as document review to thinking about e-discovery as making sense of data. Increasingly, the form of information is not limited to documents. It's audio, it's transactional, it's social, it's geo-spatial. It's requiring that e-discovery practitioners bring more of a data science orientation to their work, in order to be able to synthesize and make sense of those large volumes of data.
BH: The term data science gets used a lot. What is the distinction between just normal e-discovery and data science? Is it purely a matter of philosophy?
SR: I think it is a change in orientation from thinking about a set of documents, which is essentially just a set of particular, discrete items, that if you read each one you'll arrive at the answer, and thinking about information not as a discrete set of items but as a whole looking for trends and anomalies, and running sophisticated queries and breaking it apart into its various components, and slicing and dicing it. So it is a change in philosophy. But I think of data science as a set of skills that are required to make sense of that large, diverse volume of data. It's things like data cleansing and enrichment, not just accepting the data as it is but how you can transform it [and] make it more informative, easy to use, insightful. It requires an understanding of statistics and statistical inference, of modeling, and things like machine learning.