In August, the website TechCrunch reported that venture capital investing in legal startups had fallen precipitously from a record high in 2013 of $150 million. The headline read: “The jury is out on legal startups.”
That report generated a response article on Forbes’ website titled: “Is the legal tech boom over? It hasn’t even begun.”
If that is the case, there are plenty of technologists hovering around the legal world, looking for ways to make lawyers more efficient, more profitable or perhaps less necessary. But anyone looking to disrupt a segment of the legal economy must first answer a single question: What is it that lawyers really do?
One answer: Eat finger foods with a cocktail in hand.
Lawyers are preternatural networkers. With social calendars packed with events related to law firms, law schools, legal aid and bar groups, that’s definitely one thing they do.
So it is perhaps fitting that in November a group of lawyers started a monthly networking event in Chicago to discuss how technology is reshaping the business of law and how it can be used to augment or alter some things that (all jokes aside) they really do — review contracts, search for documents or structure law firms.
At the first Chicago Legal Innovation & Technology Meet-Up in November, more than 100 people — lawyers, professors, law students, finance pros and investors among them — went to Seyfarth Shaw to hear presentations on “the in-house view” of legal innovation and change management; how Seyfarth’s global director of legal technology innovations uses “tiny data”; what an IIT Chicago-Kent College of Law professor is doing about access to justice; and more.
The group met at McDermott Will & Emery in January to hear four more 15-minute presentations (they’re always 15 minutes, unless they are the Silicon Valley-preferred “ignite-style” presentation of six minutes) focused on technology and discovery. It was followed, in true lawyer fashion, by an “after party” at South Branch Tavern & Grill.
In February, six people spoke at the event held at IIT Chicago-Kent. Among them were Jerry Goldman and Matthew Gruhn of the law school’s Oyez Project, an online database of U.S. Supreme Court audio recordings dating back to the 1950s.
One day, they said, it might be used to analyze speech patterns that could be used to predict how individual justices will vote on a given case. In the meantime, however, they said the audio files — including those of Justice Clarence Thomas reading opinions in court — could be used to spice up songs or make pranks.
“You could take that collection of audio and chop it up in little pieces, put it back together and have (Justice Thomas) probably say anything you’d like him to say,” Goldman said. “Someone clever will discover this and take advantage of it at some point.”
The group was founded by a trio from Michigan State University School of Law: Dan Linna, Daniel Katz and Michael Bommarito. It is part of a larger trend. Similar groups exist in at least 10 other cities, including New York, Boston, Los Angeles and Washington, D.C.
“One of the things you need to do is try to create an innovation cluster around Chicago,” said Katz, an associate professor at MSU who is joining the IIT Chicago-Kent faculty this fall. “The goal of this group is to try to accelerate the pace of innovation in Chicago and in general.”
Linna, who is assistant dean of career development at MSU’s law school, said the “legal tech boom” is in an early phase where many of the young companies will fail and some will find success.
“The undeniable aspect of this is that some lawyers are figuring out how to improve their processes to be more efficient and effective,” Linna said.
“They’re figuring out how to use project management principles to get better results. They’re figuring out how to properly staff matters to be more efficient. Those things are quickly becoming undeniable truths in the marketplace. And lawyers who don’t figure out how to be more effective or efficient … are going to find it’s quickly becoming more difficult to compete with lawyers who are doing those things.”
Here are a spattering of examples of what legal tech enthusiasts have presented at the meetings so far. The next is scheduled for April 16 at Skadden Arps Slate Meagher & Flom, followed by a May 28 gathering at Chapman and Cutler.
Automated contract review
When Kingsley Martin stepped off the auditorium stage at IIT Chicago-Kent in February, an attendee at the meet-up commented that he was channeling Steve Jobs.
The remark was a nod to Martin’s black turtleneck, the look made famous by Apple’s co-founder. Martin’s presentation was also aimed, like so many of Jobs’, at making a common task simpler. In this case, the task under attack was reviewing a contract.
Martin is president and CEO of KMStandards LLC, and he was showing off an upcoming software product called Resolve. It’s ultra-simple: Drag a contract (in the form of highlighted text) into the software and tell it what kind of contract it is. Then let it go to work analyzing that contract and finding deficiencies.
KMStandards has a database of thousands of executed agreements from Fortune 500 companies that its patented software has analyzed to create “contract standards.” For instance, it can tell you what are the most important and common clauses in a non-disclosure agreement. And if you drop a non-disclosure agreement into Resolve, it recognizes what clauses are missing and which are drafted to be favorable to each party.
A law firm can also use it to analyze the standard clauses — and variability of those clauses — among the contracts it drafts.
“You can feed into it hundreds of contracts, it will tell you their standards, how they’re organized, what clauses they contain and the range of standard and non-standard terms in that set,” Martin said. “It’s the same way you learn as a lawyer, except we can process those contracts at about 100 a second.”
The program’s interface acts similarly to a Venn diagram. Terms that only exist in the contract being reviewed are highlighted in yellow. “Standard” terms in the database of contracts appear in blue. Those that appear in both are green.
The system uses exclamation marks to highlight “critical” terms — ones that it has determined are most-often negotiated, varying from contract to contract. Clicking on those terms reveals what language is used to favor each party to the contract.
On stage, Martin recounted a personal story about human contract review, what he called a nearly impossible task that his software replaces or augments.
“The first contract I had to deal with post-law school was to draft and review a credit agreement,” he said. “I had two law degrees (from Oxford University and Harvard Law School). I couldn’t do it.
“Had to do it like everybody else. Read the thing, go down the corridor, beg and borrow some credit agreements and sit down and review them. An unbelievably inefficient way to learn and an unbelievably inefficient way to practice.”
Today, he said, “nothing has changed. Literally, the partner still says to the associate, ‘Review this contract.’ And it’s left to the associate. … Hopefully, there is a better way.”
A smarter e-discovery tool
Jay Leib got laughed out of the room when he first tried selling lawyers on the idea of e-discovery.
“Lawyers love paper,” he said.
That was 2000, when Leib was in the midst of growing his first business: Advocate Solutions, which he said created the first piece of e-discovery software. Back then, he referred to it as a “processing tool.”
As he demonstrated at the Legal Tech Meet-Up at McDermott Will & Emery’s offices in January, those days are long gone.
“We’ve trained our software on hundreds of thousands of categories, and the potential for millions of categories is there. It understands internally what these categories mean. So if you start talking to it about something, it will say, ‘Oh, you’re talking about pharmaceuticals.’ It’s able to infer things.”
—NexLP founder Jay Leib
Leib sold Advocate Solutions and its Discovery Cracker product in 2003 before working for Ernst & Young and serving as chief strategy officer for kCura, the developers of the wildly popular e-discovery software Relativity.
Now, he’s back to building a business. And this time, people are cutting checks.
Leib’s NexLP raised $2 million and was picked to be one of 10 startup companies in the TechStars accelerator program in May last year. NexLP represents the second generation of e-discovery, Leib said, layering on artificial intelligence developed by Dan Roth at the University of Illinois to make human-like insights about reams of documents at warp speed. It is aimed at lawyers and white-collar criminal investigators.
The average person can review 60 documents an hour, Leib said. NexLP’s software can read six documents per second. And its speed is just the beginning. The software is designed to learn from what it reads. It can tell you, for example, whether the author of an e-mail has a positive or negative tone; whether a person of interest changed their e-mail habits; and, without any help, it can decipher the subject of a document.
“We’re essentially turning e-discovery into a big data problem and then solving it with machine learning techniques that big data companies have already been using for other industries,” Leib said.
At the meet-up in January, Leib presented along with Irina Matveeva, head of machine learning at NexLP. Matveeva, who is an adjunct professor at the Illinois Institute of Technology and, as Leib describes her, one of the “whiz kids” from Barack Obama’s re-election campaign.
She spoke about “dataless classification.” That is NexLP software’s capability to read any document and classify it. It can tell you if a document is about contracts, the law, fantasy football or automobiles, for instance.
So if a lawyer tells the software that he or she is handling a case about a particular medicine that had adverse results, the software can find the relevant documents all on its own without using keywords.
“We’ve trained our software on hundreds of thousands of categories, and the potential for millions of categories is there,” Leib said. “It understands internally what these categories mean. So if you start talking to it about something, it will say, ‘Oh, you’re talking about pharmaceuticals.’ It’s able to infer things.”
It also infers things from an e-mail user’s activity.
“What our system can do is say, ‘Hey, Jay sent out 14 e-mails over a two-week period about his largest client with a negative tone, late at night. He normally sends four. This is where we think you should look because Jay changed his behavior about this client,’” Leib said.
In 2011, the world created 1.8 zettabytes (or 1.8 trillion gigabytes) of data. That number will grow by 50 times by 2020, according to a study sponsored by data storage company EMC Corp. To sort through that massive and growing pile of data, lawyers need a new approach, Leib said.
“Technology has put lawyers, as well as investigators, in this position that there’s too much darn data and there’s not enough human beings on earth,” he said.
“There are more rules than ever before, and there’s more data than ever before. And there’s the same amount of humans. You need technology to combat that. Litigation is not going to go down. We need machines to help. And I think we’re the right team to do it.”
A partner-free law firm
In just a six-minute presentation, Brian Pike made a case for scrapping a legal establishment with a century or more of precedent — the law firm partnership.
“It’s only inevitable that you take the artisanship out of the law and put it into computers,” Pike said.
“For some people, saying you’re more like a software developer than you are an attorney — if that means sitting, researching and writing briefs — it’s a little scary.”
—Brian Pike
Pretty big ideas for a 24-year-old 3L at Michigan State. He arrived at these conclusions following a course that ended in the fall, Design Thinking for Legal Services, which he said is the first class taught at a law school focused on design thinking.
Taught by Joshua Kubicki, president of the Legal Transformation Institute, the course’s main project was to design a service delivery model for a real-life law firm to implement with a real-life insurance client. While Pike can’t share the law firm or insurance firm’s names, they were faced with a common reality in today’s legal world.
The insurance firm had previously worked with about 100 law firms. It was looking to reduce that number. Toward that end, it asked its law firms to use a Ty Metrix billing program that would allow the insurance firm to analyze and better appropriate its legal spending.
The insurance defense law firm’s goal was to earn as much work from the client as possible.
Pike’s solution was aimed at the law firm. It would create a subsidiary to work much closer with the in-house department and that would work much more efficiently — implementing Lean process management skills to eliminate redundant steps in similar legal matters.
The in-house department would first triage each matter, labeling them high complexity or low complexity. A coding department comprised of paralegals and junior associates would then apply a code to each “distinctive feature” of the case — dollar amount, counsel, plaintiffs, injuries, etc. That coding is a basis for streamlining projects in the future and can also be used to analyze the insurance firm’s legal problems, Pike said.
“Maybe in the summer, there are more slip-and-falls in one area of the state than others,” he said.
The low-complexity work would be handled by paralegals and junior associates. More senior attorneys would handle the “bespoke” high-complexity work.
Client teams — which could be based around geography or matter type — would include “workers,” “managers” and a “director” position overseeing all the work.
“Partnership wouldn’t necessarily need to be a feature,” he said, adding that some ethical rules would need to be ironed out — mainly the fact that non-lawyers cannot own law firms.
Pike’s group presented the new law firm model to the insurance firm at the end of the class. They asked if such a law firm existed, would the insurance firm hire it? They never got an answer, but Pike is convinced that more efficient legal services are the future.
That’s why he’s hoping to get a job designing legal solutions for a law firm that has positions similar to Seyfarth Shaw’s “legal solutions architects.”
“This doesn’t look like what you’d normally expect to do coming out of law school,” Pike said.
“And for some people, saying you’re more like a software developer than you are an attorney — if that means sitting, researching and writing briefs — it’s a little scary. … But I would love to sit in this consultant-type role, coming up with these deep, meaningful solutions geared toward the customer.”