Building a Better Play List with Technology-Assisted Document Review

Predictive coding can cull through vast amounts of data by quickly identifying relevant documents, thus making the document review process itself less time consuming. Machines do not get tired, lazy or distracted, but attorneys need to train the system by reviewing a sample of the overall document set that empowers the machine to learn how to rank and sort.

For parties that have yet to embrace predictive coding, there are various methods that can serve as an entry point. One of them is “document prioritization,” where the computer simply ranks documents that are likely to be responsive. Then attorneys select those documents for review while, to save costs, lower-cost reviewers are tasked with focusing on documents that the prioritization algorithm has deemed “not likely to be responsive.”

An ancillary benefit of the prioritization workflow is that it supports potential proportionality arguments. Because the documents most likely to be relevant percolate to the top of the pile, if the review process needs to be stopped it is possible to argue that the most relevant documents have been found.

Different vendors and technologies are applying  different prioritization methods. In some cases, they review the most important files first, but still review every document. In other cases, quality control and sampling metrics build strong arguments to produce documents based on predictive categorizations.

A savvy legal team will know when to leverage each scenario to avoid risk and maximize cost and time savings.