Part 1 of this post introduced some strategies for combining predictive coding with analytics technologies that were discussed on one of FTI’s LegalTech panels in February. Senior managing director Kathryn McCarthy moderated the discussion with Honorable Judge Andrew J. Peck, Jason Lichter from Pepper Hamilton and Eric Leiber of Toyota Motor Sales. In addition to covering case law around predictive coding, they offered three key use cases for when these technologies can best be applied together. In the earlier post, we summarized their recommendations for second requests. Here we discuss internal investigations and depositions.
Internal investigations can be spurred by any group inside an organization – HR, legal, accounting, etc. – or external regulators, and are a huge drain on company resources. It is a nuanced process, and once an allegation has been made, the legal and investigatory teams must start looking through data repositories quickly to garner information. The panelists offered a few examples of investigations to illustrate the deep level of detail that is often involved and why advanced technologies are necessary to get the job done. The panel suggested that keyword searching, looking at key concepts, clustering and data visualizations are all techniques that can pull the entire picture of the matter into view and begin understanding key facts, while predictive coding can be used to quickly collect and cull the evidence to meet tight timelines.
Depositions and trial prep
Analytics and predictive coding can help handle the back and forth nature of inbound and outbound data between parties during the course of a case. The panel discussed how analytics can provide a bird’s eye view of the matter before depositions begin, so attorneys have a good handle on where to focus energies. Predictive coding can be to apply the same coding to inbound data that has been decided for outbound data that is being shared with opposing counsel. This can avoid inadvertent productions and ensure that only the necessary information is being shared with the court. Judge Peck echoed his suggestions, and said that using predictive coding on inbound data during a case can be a good way to try the technology out for attorneys that are still uncomfortable using it.