LexisNexis JV Partner Knowable Launches GenAI Doc Q&A

Knowable, which has been in a deep joint-venture with LexisNexis since 2019 after spinning out of Axiom, is launching a genAI contract Q&A capability to surface insights into executed agreements held by corporates.

Nik Reed, the new CEO of Knowable, told Artificial Lawyer that the company was not a CLM business and its focus was very much on already signed agreements, not on document creation, templating, or other aspects you might find in a CLM platform. (See below for more of our interview).

The LLM-based capability, branded as ‘Ask Knowable’, taps a range of models including those of OpenAI, Anthropic, Llama, and Mistral, among others. The company will however continue to make use of several hundred human contract experts that it employs around the world.

These teams, along with earlier forms of machine learning contract analysis, have been working on the doc collections of large corporates for several years. That means that there is plenty of meta-data and access to the needed files, whether they were stored in DMSs, CLMs, or just dotted across a business.

In turn, this foundational work makes the responses of the LLMs much more useful, as a lot of the hard work around data curation has already been made for each of its corporate clients.

So, with that in mind, what can they now do with this genAI addition? Here are some key features of the Ask Knowable suite:

– ‘AI Contract Search: Locate contracts instantly using conversational prompts without complex filters or data modelling expertise.

– AI Contract Summaries: Quickly digest critical terms and obligations with AI-generated summaries of individual agreements and full contract families.

– AI Chat: Get answers to specific questions by engaging directly with advanced Generative AI that can understand the evolution of contract terms across interrelated agreements leveraging Knowable’s proprietary family mapping.’

Ask Knowable will be available to some of its clients immediately, while general release will be in March 2025.

Talking to this site, Reed also said that: ‘Our clients – which include major Fortune 500 companies – want to ask questions about their contracts. In the past we provided this with Boolean search and various filters [and via the human contract team’s input mentioned above].’

‘One of most powerful things with LLMs is the ability to ask questions, and it’s really elegant how we do this.’

Screenshot of how it works.

He added that there was huge demand for this kind of capability, with one company they know receiving over 70,000 questions a year across the business that needed to be answered by looking at the agreements it held.

Moreover, a lot of companies cannot even find their contracts and may be using a broad mix of ways to store information. This creates holes in the record and is also expensive, leading to millions of dollars in extra costs in terms of time spent by legal teams for a large business just to know what its own contracts say.

The reality is that once agreements are executed it gets messy,’ Reed added.

And that is where Knowable has come in, helping companies to get on top of these stores of business and legal data. And now they can add LLM-driven queries on top.

‘We can use the data we already have,’ he continued, ‘and the contracts have already been cleaned and de-duped. It’s pristine, and there is meta-data there too. We can see the status of contracts and what families they belong to. It’s quality controlled.’

So, when someone in a company wants to ask: how can I terminate this agreement? They can be very confident they will find the right information.

Reed concluded: ‘No-one wants to pay a lawyer to search for things. I want a lawyer to get to the point of action.’

Amen to that.