BLM: “You don’t get AI in a box”

BLM’s IT director Abby Ewen this month talked to Legal IT Insider about the UK top 40 law firm’s selection of iManage RAVN and it’s worth flagging a couple of points she made about the dos and don’ts of using AI technology.
BLM is using RAVN Extract to help capture unstructured data from its documents and emails which, working together with its data analytics team and in partnership with the London School of Economics (LSE), will form the basis of predictions around costs and claims outcomes.
BLM handles nearly 70,000 cases per year on behalf of insurance companies and currently much, if not most, of its data is held in unstructured form in documents stored across the business.
What sealed selection of the RAVN tool was a proof of concept in which RAVN was able to locate an accident report form that was located in an attachment containing multiple documents within an email box containing 7,000 emails.
Ewen told us: “We have all kinds of use cases in mind and we are forming a triangle between RAVN, our data analytics team and our partnership with LSE: we’re using RAVN to extract the data which the LSE team can then use to inform their thinking. We have loads of data, including structured data, that comes out of our case management system and this is a way to amalgamate all our data and then to build complex algorithms from which we can make predictions and build tools.”
But the bit that is worth revisiting is her observation that law firms can’t just buy complex tech and expect it to work without further input. She told us: “We often don’t have the right capability to know how to use this kind of technology and many people struggle to make it work. What is a shame is that many expect these kinds of tools to work out of the box, and when they don’t the tools get the blame.”
She added: “You don’t get AI in a box – you need the business input because the business needs to be defining the question you need to answer. What is the client problem you’re trying to solve? It has to be defined.”