Evisort Focuses on NLP Training, After ‘No Humans in the Loop’ Push

CLM 2.0 company Evisort, which has long had a strong emphasis on its ‘no humans in the loop’ NLP capabilities for contract analysis and data extraction, is launching an Automation Hub to help customers train up the system for less common clause types and unusual wording.

As the US-based company explained: ‘Automation Hub will enable anyone to easily train Evisort’s proprietary, contract-focused AI to recognise clauses unique to their business with just a handful of examples [and] no coding experience is necessary.

‘The AI allows users to easily find and report on unique provisions within their company’s contract portfolio, leading to improved workflows, accelerated deals, and faster answers to important contract questions.’

And they added in relation to ALSPs and law firms: ‘With Automation Hub, anyone can turn their contracts into insights without expensive outsourced review to track the provisions their business needs.’

This is noteworthy because the fast-growing company, which earlier this year bagged $100m in fresh funding, has often put a strong focus on how it can analyse and extract key clauses and the data within them without needing a ‘human in the loop’. This has also been marketed as their ‘self-populating dashboard’, with the onus on the phrase ‘self-populating’, i.e. you don’t need to intervene to get accurate results.

And, in walk throughs with Artificial Lawyer (see demo here) the company did indeed show what they are capable of. But, clearly what this now allows is for the system to be more adaptable. The reality is that no matter how good your pre-trained NLP is, or how many examples you have found to teach it, there will always be companies with clauses that are so unique that pre-trained NLP systems can’t always work as well as you want.

They also stressed that the training process is easy and does not need any technical skills. That said, most NLP doc review tools are also fairly easy to train and don’t need tech knowledge – but they do absorb time, and they do need someone on the team to be responsible for getting to a point where they are confident they are now getting good results.

Amine Anoun, Chief Technology Officer at Evisort, said: ‘It’s more important than ever in this economic downturn for companies to know what’s in their contracts, yet so few can do so easily today. Evisort’s goal has always been to democratise artificial intelligence. We believe you shouldn’t need a data science degree to benefit from the power of AI, and Automation Hub is built so anyone can use it without writing a line of code.’

In addition, the Hub will also offer ‘advanced search’ for the information you want to find, and there is also, as part of the new feature launch, a ‘document groups’ capability that ‘associates contracts [together] through logical relationships, making it simple to understand agreements in context of historical data, counter-parties, effective dates, and obligations’.

1 Trackback / Pingback

  1. Automated Legalese Gets More Customizable - Robot Writers AI

Comments are closed.