Reinventing the wheel

Noah Waisberg likes to remind people that while litigation document review has been revolutionized by e-discovery, corporate due diligence review is still done the old-fashioned way: Locking up law students in a room to sift through contracts, page by page, and pick out relevant clauses, to be compiled into summary charts and passed on to senior lawyers who will produce a due diligence report. And what’s wrong with that?  Well, it’s time consuming. Law students hate the work. And, according to Waisberg, “it tends to be prone to human error.” In February at ReInvent Law NYC, he explained how his product uses machine learning to make reviewing contracts faster and more comprehensive.  National caught up with the co-founder of Toronto-based DiligenceEngine.

National: So, what was the idea behind DiligenceEngine?

Noah Waisberg: Well, the idea with due diligence is that people read through the same stuff over and over again.  They’re looking at change of control or assignment provisions or something like that.  It’s a slow process; people make mistakes. And they tend to be looking for the same things over and over again.  Our idea was that we could teach software to find those same things.  So, I got together with a University of Waterloo PhD in computer science [co-founder Alexander Hudek], and we started building software to do that.  So what our software does is find contract provisions, like change of control or assignment or governing law or exclusivity provisions in a contract.  And it prepares summaries on its own.  What it really does is more enable people to gather this information more quickly. 

N: How much more quickly?

NW: We’ve done some small-scale tests. We compared me using the system with another person who graduated the same year from law school as me and worked at a large international firm after graduating. Reading page by page, I used our system and he didn’t. I got through in about 20-30 percent less time. When I just spot-checked the systems results, as a mid-level associate would supervise a junior’s work, I got through in about 60 per cent less time.  The cool part though is actually not the speed.  It’s that in both cases I was actually more accurate than the non-user. 

N: How have lawyers – whose bread and butter this is – reacted?

NW: We put a lot of effort into showing our system to large companies that do a lot of deals and they tend to be really enthusiastic.  When I show it to law firms, some are enthusiastic, but it’s more mixed.  It’s definitely not easy to sell tech efficiency to law firms. So we focus a lot more on how our software helps accuracy. Most people who work at a good law firm, and are careful, have had experience catching mistakes made by junior lawyers. So they understand how it could happen. And most law firms are uncomfortable with the idea of software doing this work on its own, without associates.  So what we’ve tried to do with the software is have them see it as a really good helper for a junior associate or an articling student.

N: You’ve also claimed that lawyers who use automated contract review can actually generate more work for themselves?

NW: Yes. What we hear of in a lot of diligence projects these days is that lawyers are getting told to only focus on the top 100 contracts.  But think of it. The average Fortune 500 company has 40,000 contracts.  And a company listed on the Toronto stock exchange, or the Montreal stock exchange, will have 5,000 maybe 10,000 contracts.  What software like ours can do is help lawyers convince their clients to do a deeper review because they can do it more efficiently.  It’s not like the clients wouldn’t like to know and it’s not like there isn’t good potential value in looking for other stuff. So automated software can make it cost effective enough that the lawyers can actually get extra work.

N: So are you a competitor to the traditional law firm, or do you play a complementary role?

NW: I can understand how a firm could see us as an adversary, but we’re just trying to help them do work that they do right now, but better.  Nearly all the clients who use our system are in large companies and large law firms – very traditional places, not at all revolutionary.

N: It’s difficult for non-traditional legal outfits such as yours to raise capital to grow in Canada and the U.S.  Any thoughts on how the legal market in England is going about things, and what it means for us?

NW: I think the scary thing though for anyone that’s trying to be innovative in Canada or in the States is that because alternative business structures are allowed in the U.K. it’s giving people there a good head start to innovate and get it right.  Look at Riverview Law. They’re working [at changing legal services] at a time when they don’t have a lot of big competitors.  So they have time to hone their model. It’s the same thing for Radiant Law [a firm that specialises in technology, outsourcing and commercial contracts].  It started with a few guys sitting around and now it’s getting bigger and bigger. And they’re starting to figure out what works. I’m not saying that Riverview is going to take over the world. But when other jurisdictions allow ABS, those U.K. firms are going to have had an advantage, I think.

This interview has been edited and condensed for publication.