Using predictive coding to accelerate document review
Discovery case studies
A global renewable energy company was in the final stages of discovery for a breach of contract litigation when they uncovered additional documents required for the discovery request. Concerned about the uncovered data volume, a looming deadline and budget pressures, the company needed help to review and produce hundreds of thousands of documents on time and within budget. The Deloitte Discovery project team leveraged Dynamic Review, Deloitte’s proprietary advanced analytics process, to accelerate the sorting and prioritization of documents for review.
The company and its outside counsel had already spent a significant amount of time and money responding to the broad discovery request. As the discovery process was nearing the production deadline, the client was faced with an additional 780,000 documents that required review. With fewer than six weeks before discovery closed, there was no guarantee that even a large review team working overtime could meet the production deadline, and the attempt would be extremely costly. Instead, the company’s outside counsel needed a cost-saving alternative.
How we helped
Deloitte assembled a team of discovery professionals to work with counsel to develop a strategy leveraging Dynamic Review. Using advanced text analytics, our professionals built a categorization model to help prioritize highly responsive documents for priority review and production. In addition, we were able to help identify potentially non-responsive documents so that counsel could allocate lower cost reviewers to this population, which contributed to overall review costs.
Once text analytics were finalized and documents were scored, our project management team and the client’s outside counsel collaborated face-to-face to develop a review strategy to meet the goals of the project. Through prioritization and review segmentation, the review was completed on time and on budget. At the projects conclusion, more than 315,000 documents were produced.
Several key factors contributed to the success of the project:
- Deloitte’s text categorization model, coupled with our experienced review team and technology-assisted review techniques, reduced the amount of documents for review
- Segmentation of the review population into distinct work streams improved workflow and throughput
- Documents scored as highly responsive with privilege were reviewed by experienced Deloitte document review attorneys reducing the level of quality control required by supervising counsel
- Documents scored as likely to be not responsive were bulk coded as not responsive and a reverse review was performed on each document to confirm the coding by Deloitte’s off-shore review team, taking advantage of lower cost review options
Leveraging Dynamic Review allowed our discovery team to assist with completing review of the 780,000 additional documents on time and within the client’s budget. When the project was completed, this approach provided approximately a 60 percent cost-saving when compared with a traditional linear review for the same volume of documents.