Site Loader

Welcome to the fourth part of our Q&A mini-series with some of the experts we interviewed in our recent Future of Legal Services Report...

EMILY FOGES, CEO, LUMINANCE

Can you tell us a little bit about Luminance and how you are working with law firms and in-house legal departments?

“Luminance is first and foremost a machine learning company. The original idea for the technology was can a computer read and understand language and learn from the interaction with humans to get better at understanding that language? So that is the starting point for everything that Luminance does. The idea was that the legal profession is where humans need the most help with reading documentation because it’s been a long time since lawyers have been able to get their heads around a corporation just by reading through a few files. When you look at processes like M&A due diligence or e-discovery, there’s way too much documentation for humans to read on their own. We came to market initially with M&A due diligence and we’ve now got over 170 customers in 40 countries, working in over 80 languages. Since then we have also developed a number of other propositions which take the same core technology and then apply different legal thought processes to that. One of them is the discovery thought process, which is much more about litigation, arbitration and disputes looking at communications between people, and then more recently we’ve started to work with in-house legal teams who are doing contract negotiation. That’s different because they’re not so much looking back at stuff that’s happened, you’re looking at live contracts that you’re negotiating right now and comparing the terms in those contracts with things that you’ve negotiated previously, and the reasons why you would negotiate particular things in different circumstances. The result of using Luminance in those situations is pretty phenomenal. We’re not talking about a 10% efficiency saving here, you’re talking about accelerating the speed at which lawyers can review documents by 100% initially, and then that accelerates to crazy numbers—one case study showed an 11,000% acceleration.”

How does your technology work?

“The first thing that Luminance does is find all of the pattens in the language. It’s not going to say this is important or that is important, it’s going to say these are the same, these bits are different. So it does that first, and then as you interact with it and you say this difference here means this or this standard wording here means that, then Luminance learns from that and builds its own rules to recognise more and more of those concepts, so it only focuses on and surfaces what you care about. It’s not something that is preconfigured to say please tell me where the risk is in this transaction, that’s impossible to do because the risk will be different depending on what the scenario is. Luminance doesn’t care—all it cares about is that you, the lawyer, apply meaning to the language, and then Luminance will say if that matches, then probably this does too, and so on.”

Given the efficiency savings, why isn’t every law firm and in-house team using your tech?

“In mature markets where you’ve gone through a few cycles of AI attempts already, they will have an understanding of technology that’s based on legacy technology, which means they look at Luminance and they don’t get it, because they have preconceptions of what it should do or how it should work. In emerging markets we don’t have that issue. So one reason is AI fatigue, organisations that have adopted something else that was less mature, and invested a lot of time and energy into it, it’s very hard to then justify doing something different. Another reason might be that law firms make a lot of money out of spending a lot of time on matters, and if you accelerate things by that amount, what are you going to charge for it? So working out the pricing models around this is definitely something that they need to consider. If you take due diligence as an example, getting the work done faster gives you an advantage because if you understand the parameters of a deal better than everybody else before anyone else, you can negotiate more powerfully. Therefore speed is good, so why are you charging more to spend longer? Why would you not have premium pricing for early closing? Because this is so transformational, a lot of the norms of the way that the legal profession works commercially are challenged, but it doesn’t necessarily mean that you charge for less hours. One of our customers said they used to spend 80% of their time being detectives and 20% being lawyers, now with Luminance they spend 20% of their time being detectives and 80% of their time doing the lawyer part, they're spending the same amount of time overall because they’re uncovering more issues that need to be advised on, so the value that you get, the more in-depth review that you get, it’s a better outcome for the client.”

How do you rate the pace at which the legal industry is adopting technology?

“There is a whole community in the legal profession—with their campaign Bring Back Boring—who are arguing that there’s no point doing anything innovative or transformational, what you need to do is reengineer your processes so you can shave 10% off here or there. That is sensible enough, but it misses the point that technology has moved on so much that you’re going to get left behind. That tells you a lot about where the legal profession is today. People need to move on from this idea that you are looking at 10% efficiency savings, because there is such transformational change coming, you need to be more ambitious, and move fast, because otherwise everybody else will get there before you. Really quickly it’s going to become very normal to use technology to help you understand lots of documentation. There was a tipping point with email, there was a tipping point with Excel where suddenly you were mad not to use it, and we are going to get to that point very soon with using AI to read documentation.”

Click here to download the full report.

Post Author: Ben Edwards