Weightmans in June announced that it has entered a partnership with the University of Liverpool and Kira Systems to develop an AI-backed decision engine to help identify arguments for settling cases – interesting in itself, there is more to the partnership than that.
In what you might call the first ‘limb’ of the project, Weightmans is using Kira’s technology to extract routine client data that needs to be entered into its case management system in an often time consuming and repetitive fashion on opening a matter.
Weightmans’ director of innovation Stuart Whittle said: “What we’re trying to do when we open a matter is to take some of the documents from our case management system and pass them to Kira, then notify the case handler that Kira has read the document and extracted the data. A human can cast their eye over whether Kira has extracted all of the data and if not, then fill in any blanks and that the information that Kira has extracted is right, and if not change it. Once that has been signed off, we’ll put the data into the CMS.”
He adds: “Most of our clients require us to input large volumes of information. We want to make the capturing of that management information less time consuming, more accurate and more consistent. We also want to make sure our lawyers’ time is spent in the best possible way.”
This is a new way of using Kira’s technology and Whittle said: “This is an unusual use case for Kira. However, it’s been used by a lot of large firms so is proven technology. Plus, their approach to the proof of concept made it really easy and cost effective for us to test whether could help us.”
The second limb of the project sees Weightmans partner with the University of Liverpool to develop a decision engine capable of assessing litigation outcomes, and Whittle says: “We want to get to the stage where we take the data from Kira and not only pass that to the CMS but also to the decision engine.”
Weightmans anticipates the technology being used by Autumn this year. Kira might ask, for example, ‘was the claimant provided with hearing protection?’ and Whittle says: “That’s a question that the decision engine needs to know.”
He adds: “We’re using Kira to extract that information from, e.g. the claim form to enable us to turn it into a yes /no answer for the decision engine.”
The decision engine has far-reaching potential and Whittle says: “My hypothesis is that if it works in this area, we can use it for quite a lot of different areas, and we’ll end up with faster, better decision making.
“It will either say ‘yes settle’ or ‘no fight’ or will tell us that if we get this and this bit of information we’ll save a significant amount of time in reaching a decision.”
This will be of most benefit in the likes of volume insurance litigation. Whittle says: “It’s not going to work for a one-off, enormous piece of litigation: that’s not the problem we’re trying to solve. It will potentially apply to anything where we have a sufficient amount of predictable and/or repeatable work.”
This story first appeared in the Orange Rag – you can read the free newsletter here: http://www.legaltechnology.com//latest-newsletter/