Catalyst Unveils Advanced Predictive Coding Tool Insight Predict

USA/Canada News

30
Jan
2013

Editor

Today at Legal Tech New York, Catalyst Repository Systems, Inc., a pioneer and long-time industry leader in cloud-based document repositories for e-discovery and other complex legal matters, announced the release of its integrated predictive coding tool for e-discovery, Insight Predict.

With Insight Predict, review teams are able to identify the most relevant documents much earlier in the review process, making the entire review workflow more focused and efficient. Catalyst is showcasing Catalyst Insight Predict at Legal Tech, Booth #2401-2403.

“Insight Predict revolutionizes the way corporate legal departments and law firms approach the entire review process,” said John Tredennick, CEO of Catalyst. “With Insight Predict, legal teams are able to run smarter searches and identify relevant documents faster, document tagging is more accurate and consistent, and attorneys can identify key legal issues and assess exposure much earlier in the process. All of this translates into more efficient, less expensive review.”

Catalyst Insight Predict seamlessly integrates into Catalyst Insight, the first cloud-based e-discovery platform to harness the power of a Big Data XML engine. 

Catalyst’s Insight Predict is based on more than four years of experience using and testing predictive coding techniques. It uses artificial intelligence to streamline and enhance human review. Insight Predict distinguishes itself from most predictive coding applications available today in several ways:  

• The technology employs contextual diversity sampling and other statistical techniques to rank each document based on a sample, or “seed set,” from a larger population of documents. Review experts review the sample and rank documents for relevance, thereby “training” the program to calculate the relevance of documents on its own.
• In an iterative, self-correcting process called Predictive RankingSM—Catalyst’s proprietary process for technology assisted review—Insight Predict feeds additional samples to the expert to test or validate their relevance and to progressively improve the accuracy and precision of its rankings.
• It is not necessary to have a complete collection of documents to create a statistically valid result. Insight Predict was designed specifically to work dynamically with rolling data input, which aligns closely with the review process in the real world, where rolling collection is the norm.

 “The great thing about Insight Predict is that you can add new documents of any type at any time,” said Tredennick. “You don’t have to wait weeks or months until all the documents in a big case are collected. And you don’t have to repeat step one of the standard workflow when those additional documents arrive.”