We’ve seen APIs and consolidation, but the future of legal tech could also be about using a component approach, or what we can call the ‘Iron Man strategy’, i.e. Jarvis (AI tech or other specialised software), the suit (the workflows and product infrastructure), plus the user.
Another way of looking at this would be to consider what automobile, fashion, and computer manufacturers have been doing for a long time via the OEM approach, or Original Equipment Manufacturer strategy.
OEM is where a number of companies may provide the components to build a computer, e.g. the main processor by Intel, the graphics card by Nvidia, the operating system by Microsoft. That computer is then sold under its own brand, for example, Dell.
The key point here is that Dell has not bought Nvidia, it has not done an M&A deal and consolidated the company as part of a platformisation strategy. Nor are Dell and Nvidia operating together via just a loose API.
In this scenario, Nvidia remains an independent company, but supplies its component product to Dell, which then makes the finished product, along with providing its own tooling, additional software, and homemade own brand parts. Importantly, Dell handles the end customer sales experience.
Yet, the buyers are not unaware that this computer has the latest Nvidia graphics card inside of it. Some may buy this particular Dell computer specifically because of its components…and also because of the better price that this OEM approach allows.
Ironclad and Iron Man
So, how does this work in the legal tech world? Let’s take a recent example: Ironclad and Google’s homegrown NLP suite for contract analysis, Contract DocAI. As explored last week, Google is taking a partnership approach where they become the ‘AI core’ that other companies will use as part of their offering to end clients.
As Jason Boehmig, CEO of CLM pioneer, Ironclad, which is the first legal tech company to partner with Google for this ‘component’, explained to Artificial Lawyer – and who is also the inspiration for this article: ‘We are trying to build the first Iron Man suit for lawyers.’
And in this case Jarvis is Contract DocAI, the suit is all the workflows, automation and infrastructure provided by Ironclad, and the wearer of the suit is whichever corporate legal team has a licence with the CLM company – and each may use it in a slightly different way according to their needs.
As Boehmig explained to Artificial Lawyer previously, Cai GoGwilt, Co-Founder and CTO of Ironclad, had looked at various approaches to using NLP to extract key data from contracts and came to the conclusion that it was simply better to wait until the right partner came along to provide what was needed, rather than trying to become a legal sector NLP specialist software manufacturer itself.
Or as Boehmig explained: ‘We want to do this for the long term, we are setting up the right infrastructure layer so we have this incredible workflow platform.’
‘Lots of fields need to benefit from the revolutionary advances in AI, that’s why we are so excited about the relationship [with Google].’
Plus, he added: ‘Lots of advances [in NLP] have not been made yet, there will be 10 more big steps and we can’t predict what they will be. [But] we are allied with the greatest hard tech problem solving company on the planet.’
I.e. this isn’t just about having a component that you need in your legal tech product right now, but also for the future changes that always arrive in the world of software.
It’s also about specialisation – which one could argue is one of the key pillars of civilisation (see Adam Smith). Why sink millions of dollars of capital into building an average component, especially when specialist developers are all over that subject and are in fact pioneering it in ways your own team will always struggle to do?
The Rest of the Legal Tech Market
Ironclad has driven this subject up to the top of the agenda, but it’s not the only company taking this approach. Zuva, the corporate-focused spin-out from Kira Systems, is taking a similar approach to Google in that they actively want to work with other tech companies in the contract and business document space.
Zuva isn’t looking to be bought outright by anyone, and some very narrow APIs are probably not going to be sufficient either. They want to work hand in hand with providers of broader solutions.
We can also see this with no-code systems. In fact, one could say that systems such as Neota Logic, AUTTO, Autologyx and Bryter are the quintessential component offerings. Many law firms have utilised the above providers to make client-facing products, adding in their own firm-specific tooling and workflows, often that have been built by the firm’s own tech team.
However, if we look into the past, we can see other situations where this could have happened, but ended up in an M&A deal instead.
For example, LexisNexis wanted to get on top of using NLP for legal research. It didn’t want to do all of this by itself so it bought two startups, LexMachina and RAVEL, to bring aboard the knowhow it wanted. It then took this, threw its tremendous resources at the challenge, and improved its existing case law platform with these NLP elements.
However, what if LexisNexis had not done that? What if companies such as LexMachina, RAVEL, or others, had said: ‘We want to work with you. But, how about we just give you a licence to our software and we become a component in your offering?’
How far can this approach go? It feels like we are only just starting to really explore the full potential of this component, or ‘legal tech OEM’, strategy as a sector.
And here is one other example of where things could change. Take doc automation. There are many providers of doc automation solutions. And there are many legal tech products that offer this as part of their platform.
Why not have a handful of doc automation providers sell their ‘component’ capability under licence to a bunch of other companies all over the world?
How about the market decides which are the top three or four doc automation systems for legal work and then every platform that uses that approach as part of its offering then goes to one of these to install that component, rather than trying to cobble together its own, and remake the wheel, or instead feel the need to go on an M&A hunt to own one of them? And this would be more than a distant integration, this would be the doc automation system fully ‘built in’ within the larger provider. Not an add-on, not a distant API, but totally integrated within the product by design.
In fact, is this the antidote to legal tech M&A? You don’t need to buy your way to building a legal tech platform if you take the component approach.
Then, rather than a bunch of companies that don’t always fit easily together, being squeezed onto one platform, the larger companies can cut component deals with the leading providers of all the bits they are missing? Just a thought.
It will be very interesting to see who else takes this approach.
Service-oriented architecture is such a cost-effective way forward, and more importantly allows you to fail fast, and improve for rapid growth. AWS were pioneers of this, and it is quickly becoming the norm.
Smart choices all round!