Technology Assisted Review

Technology
Assisted
Review

Partner with TLS's expert discovery consultants—including highly experienced project managers and former litigators—to develop workflows utilizing machine learning technology. By surfacing relevant content faster and cutting the number of documents in need of review, TLS technology assisted review experts help you meet stringent deadlines while saving you significant time and money related to document review.

Case
Study

The Client

An AmLaw 10 law firm representing a global entertainment company that received a second request from the U.S. Department of Justice (DOJ) in regards to an acquisition valued at $60 billion.

The Challenge

Over 30 TBs of data collected in response to the second request

60 days to complete the collection, processing, review, and production of responsive data

Unusually high volume of duplicative data due to uncommonly frequent backups by the client's cloud-based backup system

The TLS Solution

Faced with these difficult circumstances, TLS, outside counsel, and the corporate client rapidly deployed a highly customized workflow consisting of three major elements: (1) forensic deduplication prior to ingestion into the e-discovery processing tool, (2) application of Pre-Review Analytics® in close consultation with the DOJ, and (3) negotiation and implementation of Technology-Assisted Review (TAR) protocols.

In the end, the DOJ agreed that its demands had been met and the client was able to meet their 60-day deadline.
 

Want more details?
Speak with a TLS legal expert who worked on the project.

"It is now black letter law that where the producing party
wants to utilize TAR for document review, courts will permit it."

Magistrate Judge Peck, Rio Tinto PLLC v Vale S.A.

(S.D.N.Y. March 3, 2015)

Predictive Coding | TAR 1.0

Establish TAR
Success Metrics

(Recall Precision, Confidence Level)
Subject Matter Expert (SME) Reviews Control Set
SME Reviews
Training Sets
Report & Validate
Results
Propagate Coding Decisions Against Remaining Documents
Produce Documents Coded Responsive After Privilege Screen
Review Documents Identified as Responsive & Then Produce

Continuous Active Learning | TAR 2.0

Review Sample Documents
Update Predictive Model
Achieve Point of Diminished Returns*

*Very few potentially responsive documents remain

Sampling for Validation

Production*

*Certain reviews may have a different end goal, e.g. supporting an investigation or understanding a document dump

The Global Solution for Legal Support

We Know How