We deploy advanced analytics, such as AI and Technology Assisted Review (TAR), to analyze an entire pool of data, applying technology, workflow and subject matter understanding. These analytics allow us to reduce the data set and to segregate and organize the remaining data to get to the relevant information faster.
We also use predictive modeling in a variety of ways, including to define a review population, to approach a fixed population strategically, and conduct second-level quality control. We have built models on all major platforms that our clients often use as their preferred default approach to finding relevant content. Our approach is based on the concept of continuous active learning and combines subject matter and technical expert collaboration, supporting process documentation, and tracking of model refinement, to obtain a defensible result.