Today, we are announcing our investment into LawGeex, a company on a mission to “help businesses run faster and better by automatically answering one simple question: ‘Can I sign this?’”.
To provide context about the investment, we need to go back seven years. In August of 2011, Marc Andreesen said “software is eating the world.” Now, in 2018, this simple statement seems obvious. When we raised the first Aleph fund in 2013, we told our investors that we believed we were at the beginning of the big data, machine learning, AI era and we wanted to invest in the ability of those technologies to disrupt every industry on the planet. So, we must now take it a step further and say that today Artificial Intelligence is augmenting and replacing every profession.
In short, data and algorithms that act on that data are disrupting almost any corporation without a deep-learning plan and it’s going to get ugly for the unprepared. It will be a huge competitive advantage for those willing to take the leap
Where will AI find purpose next? I find that there are two, as of yet under-served, fields that collect information in a form suitable for machine learning:
- Regulated industries like Insurance, Real-estate, Maritime, or Transportation, since regulations require more structured data in order for them to be able to track compliance.
- Process-oriented professions, such as architects, lawyers, or accountants, who use documents and other forms of semi-structured data as a medium of exchange of work products.
LawGeex is focusing on the legal market. The legal industry makes over $400 billion annually in the US alone but spends less than one percent of that on enabling technology. If you compare the legal industry to the tech investments made by the financial industry, you will understand the vast opportunity. JPMorgan’s 2017 revenue was $94 billion, and they spent an astounding $9.5 billion on technology. That is more than 10 percent of their revenue and over ten times more spend on tech than the legal industry. We are betting this chasm is not sustainable.
We have been intrigued by tech opportunities serving or disrupting regulated and traditional industries, backing companies like WeWork, Lemonade, Windward, Freightos and Colu. This is especially true for service-providing departments within large enterprises. They are constantly being told “do more with less.” Finally, they can. This is all thanks to the little thing called annotated or tagged data, when combined with the recent advances in machine learning.
Machine learning is already in play. In recent years, directly annotated datasets like tagged images or radiology data combined with electronic medical records, have been used to teach the AI engines how to automate facial recognition and medical diagnosis, respectively.
However, indirectly annotated data is less obvious. Legal documents present a significant cache of semi-structured data that can’t be a direct input to machine learning algorithms as they don’t benefit from simple pattern matching. They require a higher-level conceptual representation. The LawGeex team is using a combination of natural language processing and deep learning to automate the analysis of legal agreements, providing actionable recommendations and insight to the business.
So, what makes them so innovative? As many of the blog readers may know, legalese is a kind of “language.” There is some order to this madness; many of the agreements “speak” about the same topics, with different nuances. This implicit structure is what allows LawGeex to be already better than human lawyers for reviewing some documents including the NDAs that were the company’s original effort.
Contrary to the slightly clickbaity title, what excited me most when talking to Noory was his belief LawGeex is the “Iron Man Exoskeleton” for lawyers. This power suit, if you will, allows them to streamline mundane tasks. Lawyers can now act as a partner and an enabler to the business. They are able to use the time they’d otherwise lose to provide a real advantage to the company. This approach is what I — a venture capitalist and a business user — have found the most appealing.
LawGeex has been on my radar for a while now, ever since they participated in our Bot Challenge. Our initial interaction was back when they were targeting their groundbreaking tech to the wrong audience: reviewing NDAs and rental agreements for consumers. Noory Bechor, the company CEO and a lawyer himself, has kept in touch throughout, pivoted their focus to serving businesses and enterprises and has since closed eBay, Farmers Insurance, and others as customers.
I am excited to finally join Noory, Ilan, the team on the journey to empower lawyers and companies. Undoubtedly, this is a pivotal point in time in the legal industry as well as for LawGeex. As one of the reference customers told me during due diligence — “Legal AI is like teenage sex: Everyone talks about it; nobody really knows how to do it.”. It is a good thing LawGeex has come of age.