Technology-assisted review (TAR)
As data volumes and pricing pressures surge, so does the adoption of cutting-edge technology and tools to make document reviews more efficient. The use of predictive coding offers law firms and corporate clients a practical strategy to streamline the review lifecycle.
Predictive Coding Strategies
Within the context of a technology-assisted review, Compliance applies predictive coding alongside our proprietary workflow and methodology to maximize results through an iterative process and quality assurance protocol. Our measured and hands-on approach reduces review time and costs and includes several clear steps:
Step 1: Alignment & training
A subject matter expert undertakes a thorough review of the case and coding rules in order to begin applying decisions to the data set, thereby increasing accuracy and consistency throughout the process.
Step 2: Collaborative seed set review
Experienced attorneys collaborate in order to reach consensus on the coding and apply decisions to an initial subset of documents, which are then applied to the entire data set and throughout the process as a quality control benchmark.
Step 3: Expert analysis
Senior attorneys review application of the initial data set coding, and inconsistencies are documented and revised. This iterative process occurs until precision and recall rates reach statistically sound ratios to move to the validation process.
Step 4: Validation and QC
Statistical sampling and detailed quality control measures are employed as part of a rigorous validation process. Refined over the course of the project, this approach provides strong defensibility and accurate results.