Situation
An international technology manufacturer faced court-ordered remediation following a trade secrets litigation matter. The company hired employees from a competitor, and those employees inadvertently brought trade secret information with them. Over time, that information spread across the client’s network through email, shared drives, and other internal systems.
When the competitor initiated litigation, the court ordered the client to identify and remove all trade secret data from its environment. An initial remediation effort by another provider relied primarily on keyword searches, which the court later deemed insufficient.
Through close collaboration, TCDI’s teams fully processed 13TB of data and developed remediation libraries within
3 weeks
As a result of the failed remediation and ongoing litigation costs, the client faced significant financial pressure and planned to sell the business. It could not move forward, however, until it demonstrated to the court and the opposing party that all trade secrets had been fully identified and removed.
To meet the court’s requirements and address the scope of the data involved, the client engaged Stroz Friedberg, a LevelBlue company, to lead a subsequent forensic effort focused on remediation. Stroz partnered with TCDI to lead trade secret identification, analytics, and reporting. These reports served as the remediation library to support the ultimate remediation of the data, satisfying the orders of the court and allowing the dispute to come to a resolution.
Resolution
The client’s environment contained approximately 33TB of data. Following scoping discussions and negotiations, the dataset was narrowed to approximately 13TB required for ongoing business operations. This scoped dataset was collected by Stroz Friedberg and provided to TCDI for processing and review.
During processing, the dataset expanded to approximately 22TB of data, representing roughly 41 million documents. Given the compressed timeline, TCDI developed a custom, streamlined workflow using the guidelines for processing and hosting very large data sets in Relativity to enable mass handling at scale.
Using a seed set of documents from a prior injunction, TCDI then applied multiple layers of analytics to the dataset to identify content closely aligned with known trade secrets. In the first pass, similarity and categorization analytics were used to surface documents most comparable to the seed set.
Results from this first pass were used to train a Continuous Active Learning (CAL) model in Relativity’s Review Center. This second phase was supplemented with competitor-provided descriptions of the trade secrets in question, publicly available reference materials, and tailored search terms developed by TCDI. Documents surfaced through this process were then reviewed by TCDI’s Military Spouse Managed Review (MSMR) team.
At the same time, TCDI’s MSMR team conducted a parallel review of image-heavy file types to identify visual materials that could represent trade secret content. Findings from both review streams were refined through collaboration with the client’s engineers to validate the classifications.
TCDI’s Systems Operations team further supported the engagement by optimizing indexing to support high-volume searching and designing custom remediation reports. Separate reports were created for individual machines as well as a consolidated enterprise-wide document detailing all the companies’ servers required for ongoing business activities. Stroz Friedberg used these reports to execute on-site remediation activities for the entire company’s network.
As the client continued operating during the engagement, TCDI also performed a gap analysis to account for data created after initial collections, resulting in a second round of collection, analysis, and remediation using the same validated process.
During the matter, TCDI provided expert witness testimony and participated in two rounds of depositions to explain the methodology, validate the remediation results, and demonstrate the ability to handle the data volume within the accelerated timeframe.
Impact
TCDI executed large-scale processing and analytics across a complex enterprise environment using coordinated, around-the-clock workflows and custom processing solutions.
Through close alignment between TCDI’s Client Services, Systems Operations, and MSMR teams, TCDI was able to establish remediation libraries within three weeks. This was accomplished by fully processing the scoped dataset of approximately 13TB (22TB expanded), establishing a cutoff rank that satisfied finding trade secret documents, and reviewing the results using CAL.
After applying layered analytics and review workflows to the 41 million documents, the dataset was reduced to approximately 400,000 to 500,000 documents available for human review, representing less than 2 percent of the original quantity.
The entire engagement, spanning initial collections through final reporting and gap analysis, was completed within a matter of months while the client continued normal business operations. At project conclusion, the client demonstrated full compliance with the court order, successfully removing all identified trade secrets from its network, and was able to proceed forward with the sale of the business.
