I have interviewed Kelly Atherton, Director of Analytics and Managed Review at NightOwl Global, several times over the years, mainly about the use of Relativity’s analytics tools. She speaks with the authority of one who uses these tools every day, and the interviews act as a record of the development of technology-assisted review.
NightOwl Global has always shown a deep commitment to the use of new software tools as they develop. I last interviewed Kelly Atherton when she and NightOwl were first starting to use Relativity’s Active Learning, and I wanted to find out how that use had developed.
Kerry Atherton began with an emphatic statement that Relativity’s Active Learning works. NightOwl had been able to compare it with past cases where TAR 1.0 had been used and, she said, always reviewed fewer documents with the new tools. You need a long history of past cases to be able to make comparisons like this in any meaningful way.
The passage of time has also given NightOwl the opportunity to adapt their workflows. If humans were now agreeing with the results of the new tools 90% of the time, a different approach may be needed to doing the job and checking it.
It is important, Kelly Atherton said, to realise that Active Learning is just one piece in the workflow. It remains important to get an idea in advance of how many responsive documents are expected, not least for budgeting purposes. You still need to anticipate the proportion of documents which don’t work well with active review such as spreadsheets and videos. Overall, she said, all the basic principles which apply with the conventional approach remain important. You are not changing your fundamental approach but getting through the work faster.
Clients are more knowledgeable than they used to be. They no longer worry, for example, about whether the technology is defensible because that is by now well established. What matters is that Relativity’s Active Learning has worked in every project, substantially reducing the number of documents for review.