DSi First to Use Catalyst Insight Predict

Global News

28
Aug
2013

Editor

Unique Technology Assisted Review Tool Reduces Review Cost by 90 Percent

Electronic discovery company DSi (formerly Document Solutions Inc.) recently became the first to use Catalyst’s predictive ranking product Insight Predict. Working with a law firm client in a case involving a government inquiry of a national healthcare company, DSi concluded that by using Insight Predict in conjunction with DSi’s proprietary sampling process, they were able to reduce workload by 72 percent and cut costs by 90 percent over traditional linear review.

Insight Predict is a Technology Assisted Review (TAR) product that uses mathematical and statistical techniques to organize large document populations by likely relevance. In this way, Insight Predict helps review teams identify the most relevant documents much earlier in the review process, making the entire review workflow more focused and efficient. 

"Insight Predict has revolutionized Technology Assisted Review for corporate legal departments and law firms by introducing the first product that works with rolling collections and real-world workflows," said John Tredennick, founder and CEO of Catalyst. “We are thrilled that DSi experienced such great success with Insight Predict in this case and we look forward to supporting its efforts to deliver a smarter, more efficient review process to clients."

Early in this case, DSi and the law firm realized that it would involve far more data than originally estimated. To cull the data down to manageable size, they started by using keyword searches based on DSi’s proprietary iterative search and sampling techniques. They tested the searches using Catalyst Insight’s PowerSearch utility and then refined the key terms to optimize results.

As a result of this process of search and sampling, the team was able to reduce the initial document set by more than 90 percent. Even with this reduction, however, the team still faced a costly and time-consuming linear review for the remaining volume of data. 

At this point, DSi suggested that the firm could further reduce the review volume by using Insight Predict. They began with an initial seed set of 300 documents and then, based on the initial results, added initial seed documents. After this second round of sampling, the review team was able to determine that it would find 80 percent of the responsive documents after reviewing just 8 percent of the total document population. In comparison, linear review would have found 80 percent of the responsive documents only after reviewing 80 percent of the total.

“Insight Predict is a very powerful tool that could revolutionize Technology Assisted Review,” said John Burchfield, DSi vice president of business development. “The results from our first case using Predict were stellar, especially compared to a manual review. When Predict was fully trained, it reduced the amount of documents to be reviewed by 72 percent compared to a linear review. This will lead to dramatic cost and time savings during attorney review, which will greatly benefit our clients.”

“We are constantly adding to our suite of tools for Technology Assisted Review, and Insight Predict is a fast, accurate tool that we are proud to offer,” said Tom Turner, president and co-founder of DSi. “We started using Predict in another case already and we are planning future uses.”