Ashurst LLP, a leading international law firm with nearly 2,000 lawyers
worldwide, represented a financial services company in a lawsuit alleging
negligence by its auditing firm. In 2013 , Ashurst enlisted Recommind
to provide an eDiscovery platform and managed service to support a
large-scale document review. In parallel, Ashurst engaged Integreon to
provide a team of legal reviewers that, at peak, totalled 80 people.
The firm sought both to keep costs down and to identify key documents
quickly for case preparation, goals that were challenged by the sheer
volume of the data set and the fact that data delivery to Ashurst and
Recommind was staggered over several months. A pressing timetable
of staggered disclosures necessitated Ashurst starting to review documents
in a linear manner. However, it was unlikely the ultimate deadline
could cost-effectively be met by linear review alone. The Ashurst
team consulted Recommind for advice about using advanced analytics
and, in particular, Predictive Coding, to reach their goals and disclosure
Leveraging the Axcelerate platform, Recommind’s managed services team developed two parallel workflows.
Under one, the Integreon document review team, supervised by Ashurst’s attorneys, started reviewing
documents employing their efficient and streamlined review workflows. Under the other, the results of that
review were used to train Axcelerate’s patented Predictive Coding technology and initiate a concurrent, analytics-
driven, prioritised review.
Under the second workflow, Axcelerate’s continuous machine learning suggested likely-relevant documents
for priority review by Integreon’s team. With each iteration, the results of human review were used to refine
the Predictive Coding engine’s training, improving the relevance of the next suggested document set, for
ever-increasing reviewer efficiency. As the presence of relevant documents in the unreviewed set fell to
near zero, it became apparent that human review of the remaining documents would be a waste of time,
effort, and money.
Throughout the review, Recommind and Integreon’s data science and project management experts guided
the Ashurst legal team with regard to technology, best practices, and defensibility. When the opposing
parties posed specific challenges, Recommind’s consultants supported the solicitors and barristers as they
represented to the court the efficacy of utilising Predictive Coding to honour disclosure deadlines while
keeping costs in check.