Within the eDiscovery software Relativity, users can review a wide array of file types. One particular file type, transcripts, is the transcribed text of a deposition. Often times, paralegals and attorneys will want to call attention to a portion of text that matches a particular theme. These are called Designations. Terms like Plaintiff, Defendant, and Joint are common.
A legal team will frequently receive an exported designations csv file from opposing counsel or other team members for one or multiple transcripts. The problem was finding a way to add those into Relativity to see a holistic view of all designations for a matter.
Through customer interviews across personas, we learned that there's a pretty common structure to the CSVs exchanged between parties. There are approximately 6 columns that are necessary to map to fields within the software, but the number of unique transcripts could vary.
We decided the best solution was to leverage the software's existing file upload user interface, while adding a wizard for the additional steps for system field and column mapping, as well as matching the transcript titles referenced in the CSV to those that are in Relativity.
Through user testing, we found that occasionally the CSVs to be uploaded have missing or incomplete data. Since this data can be pretty sensitive, we wanted to allow the user to review errors before the upload started. We decided to use a new warning style that was closer to the main calls to action in the modal when errors were found, either in the file, or during field mapping.