Interviewer: Wendy, you now use artificial intelligence to perform a number of different tasks. But before you started using it how were you doing things like data extraction and document reviews?
Wendy: Well we were doing them in what you would think of as a very traditional lawyer’s way. So we would get the documents in from a third party or our client, assess how many there were and then depending on the number we would put together a team of mainly junior lawyers, maybe paralegals as well, and they would be put in a room with the documents and either a dictaphone or a computer and told to analyse those documents for whatever data we needed, they would then usually put that data into spreadsheets or reports and hand it to the more senior lawyers. And that process could often take days or maybe weeks, depending on the exercise.
Interviewer: And Wendy, how does the artificial intelligence solution that you’ve adopted work?
Wendy: Well, we’ve decided to call our robot Lonald so I will continue to refer to it as Lonald. We have a software portal; we take the documents that we need to review; we usually turn them into pdfs; put them in zip files and upload them to the portal. And then we simply press a button which says extract data and Lonald does all the data extraction within a matter of seconds and then puts the outputs into a spreadsheet. Now that spreadsheet is something that we ourselves will have created in a bespoke manner for each case and what we will have done is created that spreadsheet which has fields in it which are effectively the questions that we’ve asked Lonald to answer. And that obviously takes a little time because that varies from case to case. In our old method what we found was that we would have issues around fatigue for example. Our staff would be working for long hours doing the data extraction and human error will always creep in in those situations and also people’s morale will suffer if they’re doing this for weeks on end. Again the robot doesn’t have a morale issue. The other thing we found was that the human beings who were doing the task would either make errors or even if they were accurate they would nonetheless not all record the data in the same way. And therefore what we would find is that we didn’t have a standard product at the end of the day and some time would have to be spent standardising it. Well again when we asked the robot to do a job it does exactly the job we’ve asked it to do.
Interviewer: Now that you’ve been using it for a while, what do you think the key advantages are?
Wendy: I think it’s worked very well for us and the main benefits have been efficiency and accuracy I would say. We have seriously reduced the amount of time that we need to spend on these activities. We’ve reduced the number of people who need to spend time on it and therefore we’ve freed up our junior lawyers to do more interesting work. I think that’s really helped with morale as well in our teams and as far as accuracy is concerned, we now have a product that we can trust to do exactly the tasks that we ask it to do and not deviate from them.
Interviewer: So Wendy this is leading to new ways of working.
Wendy: Well I suppose it is because for example we can now say to a client who’s got a very urgent piece of work or transaction to do that we can undertake certain phases of it in a very short period of time compared to saying to them this will take us two weeks because we don’t have to scale up our staff and then having scaled up the staff they don’t have to spend the time on the activities. So I think it is a new way of working for us yes.
Interviewer: So Wendy where next for this technology?
Wendy: Well I think in this age of computing we’re seeing advances the whole time and what we have learnt from our artificial intelligence providers is that the robot for want of a better word itself is learning the whole time and they are developing ways of using it in ever more sophisticated ways. So I think we will soon see a time where for a range of different documents we can teach the robot to read the document in the same way that a human would and actually give us an analysis of the main points of interest in the document. Basically it will tell us whatever we ask it to tell us. And I think if that happens it means that the lawyers themselves don’t have to act in a robotic way of producing spreadsheets of data. We can go straight into analysis and device. And that I think can only be good for our client relationships and from a client perspective it can only be good because it’s a more efficient and cheaper way of working.
Interviewer: Wendy, thank you.