Reduce Time with Technology

Computer-assisted (or computer-informed) review offers an iterative learning process to define relevancy, while coding for documents is completed by specially-trained professionals (or your software of choice when working under time limitations), providing a cross-measure of quality control.

• Hudson works with the case team to create a set of instructions based on substantive details
• Initial sampling of documents reviewed to provide basis for computer-suggested categorization and relevance
  • Small number of specialized Hudson attorneys and Project Managers conduct early assessment of the sample
• After initial sample reviewed, remaining documents are prioritized by relevance; documents can also be clustered by relevance to different topics
  • Advanced relevance-ranking technology learns from previous coding and informs decisions as review progresses
• As reviewers go through the documents, they see computer-suggested coding and its confidence level in the choice, helping inform their decisions and dramatically speeding the review
  • System learns as more documents are coded, refines future suggestions
• For clients with heavy time constraints, we can further leverage technology to achieve significant time savings by deploying cutting-edge algorithms to code the data set in a fraction of the time needed for traditional linear review

Hudson Legal Quality Control procedures include:

• Comparison of computer-suggested tagging against tags applied by review attorneys
• Results testing and adjusting computer rules iteratively until results meet acceptable accuracy levels (typically less than four iterations needed)
• Identification of documents where computer is unable to make a decision, adjustment of rules to include these documents, or rapid manual coding where needed (usually less than 10% of documents)
• Hudson's team of specialized review attorneys manually review a statistical sample of the documents to verify accuracy and ensure defensibility