Situation

Our client was working on a large group litigation matter and was tasked with reviewing various disclosure tranches received from multiple defendants. Each disclosure set involved hundreds of thousands of documents that needed to be reviewed by the legal team in a relatively tight time frame.

Further, there were both internal and external requirements for regular reporting on the progress of each disclosure review workflow. The legal team needed to find a way to undertake a quick, efficient review and ensure each review stream's progress was quickly and accurately reportable to various stakeholders.

Solution

Due to the volume of the disclosure reviews, the Sky Discovery team suggested concurrent Continuous Active Learning workflows for each disclosure set. This meant that the CAL models would assist the team in more quickly identifying relevant documents than adopting a more traditional, linear review workflow.

Further, the ongoing results of the CAL-powered review could be exported to a custom excel spreadsheet that could provide various insights on the progress and accuracy of the review to multiple stakeholders. These insights included CAL review figures, percentage relevance per CAL and count for first, QA and second pass review outcomes.

The spreadsheets were complex and detailed and initially took too much time to prepare. To limit the time and costs incurred, Sky Discovery’s technical team developed a custom reporting tool that removes large volumes of the manual work required to prepare these reports.

Impact

Our work in developing the CAL workflow and reporting mechanism automation allowed our client to:

• Schedule regular and frequent reporting intervals keeping all relevant stakeholders up to date with accurate and detailed review information

• Manage the expectations of various stakeholders as to timing, productivity, and accuracy of the review workflow

• Significantly reduced the time and fees associated with manual preparation of the reporting output.