In accordance with accepted Privacy frameworks and regulations, this website stores cookies on your computer. These cookies are used to collect information about how you interact with Lighthouse’s website and allow us to better inform and communicate about the topics you actually care about. Lighthouse uses this information in order to improve your experience and for analytics and metrics. View Privacy Policy

Coronavirus Preparedness At Lighthouse


ReviewSmart, Lighthouse’s technology-assisted review offering, incorporates multiple software solutions and unmatched project management capability. ReviewSmart is the least costly, fastest, and most accurate way to conduct first-pass review.

ReviewSmart helps reduce document populations based on an expert reviewer’s coding of a small sample of the documents. TAR eliminates first-pass review, minimizing review of non-responsive documents. This results in substantial cost savings in hosting expenses and, more importantly, attorney review hours. It also increases consistency of review decisions across your case.

How the defensible workflow works

A successful TAR depends on both technology and process. Our experts use Equivio, kCura and/or Lighthouse software, along with Lighthouse middleware and project workflows.

Lighthouse-customized workflows follow these steps:

  1. The tool generates a seed set to train the system.
  2. Expert reviewers conduct iterative review rounds to further train the system and validate its performance.
  3. Lighthouse experts assess overturn rates for each round to determine when the system has reached optimum performance and needs no additional reviewer input.
  4. Once trained and validated, the system classifies every document. Typically, the designations are “likely responsive,” “likely non-responsive,” and “uncategorized.”
  5. Attorneys review samples of documents in each category as a quality check.

With the results of the TAR process, our clients can then prioritize which documents to review first. They might decide to review only the documents the system categorizes as responsive, or simply produce the documents the system categorizes as responsive.

TAR categorizes data more accurately and quickly than human-eye linear review. By amplifying the calls of one or only a few reviewers on a small subset of documents, a precise and reduced data set for linear review is reached in days.

Data validation

A successful TAR process requires considerable statistical analysis, along with a deep understanding of the underlying technology. Parties must also be able to explain the process and decisions made at each step to opposing parties, the courts, and governmental agencies.

At Lighthouse, our process includes sample selection, validation rounds, statistical analysis, and stringent quality control. Our experts hold degrees in related fields and have decades of eDiscovery review experience. Because of our expertise and the rigorous technology in our TAR solution, our clients can be confident of a successful and defensible outcome.