The value of technology assisted review: case study

The challenge

We managed a large funds management dispute disclosure exercise, which spanned 10 years and included dozens of employees’ email accounts and Office files. With only 6 weeks to provide disclosure and over 1.7 million documents potentially responsive, could this exercise be completed?

Our solution

Continuous Active Learning (CAL), Relativity’s machine learning application, was the right approach to manage the size of the data and meet the deadline. By applying CAL to the pool of 1.7 million documents, we were able to determine that only 180,000 were potentially responsive and required review. Our market-leading eDiscovery team was able to utilise other analytics within Relativity, including email threading and near duplicate identification, to identify tranches of relevant documents that could be quickly disclosed.

The outcome

By leveraging Relativity’s machine learning capabilities and data analytics, we were able to reduce the review population by 90%, allowing us to disclose the relevant documents on time and saving our clients over £1 million in costs. We agreed to provide monthly, rolling disclosures to the other side, and were ready to produce the first disclosure tranche 8 days ahead of the scheduled deadline. By continuing with the CAL review, we were able to continue to provide rolling tranches of data for each monthly disclosure and meet the deadlines.