Knowable has now formally launched its contract data analysis platform which utilises NLP, along with some additional human review (AKA ‘lawyers-in-the loop’), to provide corporates with ongoing legal insights and business intelligence derived from their contract stacks.
The US-based company, which was once part of Axiom and is now in a JV with LexisNexis, (and featured in Artificial Lawyer in December and October, last year), has been working on the product for some time. It has tapped several years of prior contract analysis experience, initially with a mainly human review approach back when it was part of Axiom, which it has leveraged for training up machine learning software to identify and extract key legal and business data. This is then presented in a user-friendly dashboard that gives insights into the business (see below).
While their approach has been looked at before by Artificial Lawyer (see above) the company is now formally going to market with its NLP-driven system.
This will enable companies to quickly search, find and analyze contract data – whether this is derived from an ongoing doc stack, or via an incoming set of documents following an M&A deal.
Whichever way the data in on-boarded, the goal is the same: to turn unstructured text into meaningful insights that will help a corporate to run more efficiently, experience less risk, and hopefully become more profitable by not leaving beneficial contractual terms locked away and unread.
For example, KPMG estimates that 40% of the value in contracts is lost. This is because clauses that may provide additional income are simply lost to human perception as contracts get filed away and forgotten about.
Speaking to Artificial Lawyer in New York during LegalWeek, where the company won the the LegalLaunch: Product Innovation Competition, Nik Reed, who recently joined from LexisNexis and is now Senior Vice President of Product, Design and R&D at Knowable, said: ‘Reading every word in every contract may not always make sense. But, also thinking that an algorithm alone can do all the work is also not the best approach.’
He stressed that you need ‘lawyers in the loop’ to reach the very high levels of accuracy needed for their solution. While lower accuracy levels may be sufficient for a quick due diligence exercise, if you’re seeking to isolate specific terms that may relate to significant financial agreements then you need very high accuracy. And for that you need some human input as well.
‘It’s really about workflows, and machine learning is all about improving those workflows,’ Reed added.
‘Contracts are wildly undermanaged and under-leveraged today,’ said Mark Harris in an additional statement, CEO of Knowable and one of the founders of Axiom. ‘Contracts govern every aspect of every commercial arrangement a company has, making them one of the most valuable and powerful information assets in business. And yet most companies can’t even find their contracts, let alone tell you what’s in them.’
The company is now aiming to increase its client base, which already includes several Fortune 500 corporations.