An international law firm required eDiscovery services to assist its client with a legal dispute based in a foreign jurisdiction. The client had approximately 1,000GB of email data and eFiles, electronically stored files, that required processing, review and disclosure.
The law firm was not sure where to start with such a large pool of data, but was having to work to an extremely tight, three-week turnaround time to complete the initial investigation phase.
After discussing the requirements with the law firm, our forensic technology team collected and processed close to three million documents for review in the firm’s document review platform, Relativity. Based on experience in similar cases, the team estimated that around 10% of the documents would respond to keyword searches, which would generate a review set of around 300,000 documents.
The firm had considered using 10 reviewers, but even such a large team would have needed more than two months to review the documents. Following further consultation, our team introduced the use of Continuous Active Learning (CAL).
CAL identifies and prioritises documents for review based on the input it receives from the reviewers by ranking documents according to how likely they are to be responsive and moving the highest ranking documents to the front of the review queue. Our team were able to train the client on how to get the best out of CAL – a relatively new technology for them – and as a result were able to reduce the size of the dataset from three million documents to 25,000. A quality control review resulted in a very low overturn rate, with only 0.2% of the documents that had been classified as not relevant by CAL subsequently found to be relevant.
The law firm completed the initial review in under a week, saving an estimated £1.25 million in potential review costs. After reviewing the 25,000 documents that CAL flagged as likely relevant, it then disclosed around 1,000 documents to the opposing solicitor.
As a result of this work, our forensic technology team are now a preferred supplier of the firm for eDiscovery services.
The estimated savings in review costs through leveraging Continuous Active Learning (CAL)