Supporting a Data Driven Look at AI Accuracy in Document Review

Cimplifi was proud to support the Relativity Redgrave Accuracy Study as a neutral participant, helping establish a real‑world baseline for evaluating modern generative AI‑enabled review workflows.

The study was designed to compare traditional first‑level review to generative AI‑driven approaches using long‑established, court‑accepted validation methods. This study validates that generative AI document review (aiR for Review) is a defensible, empirically validated alternative to traditional managed review under established legal standards like the Sedona Conference TAR 1 Reference Model and Rule 26(g). The study also highlights that practitioners can confidently adopt GenAI review as a primary classifier, second level review, or for expert augmentation.

The role of Cimplifi was intentional and controlled to preserve the integrity of the findings:

• We hosted the data in a neutral capacity
• We conducted a standard first‑level managed review using Relativity Active Learning (RAL) under the supervision of the Redgrave team.
• Our review team performed the work without knowing it was part of a study, ensuring unbiased results

The findings reinforce a powerful shift in legal review: AI‑enabled workflows can deliver higher accuracy, reach defensible production points faster, and allow legal teams to focus effort where it matters most.

The study reflects a baseline comparison. In real engagements, Cimplifi builds on this foundation by pairing advanced technology with quality control, validation, and expert oversight to deliver defensible outcomes clients can trust.

For Cimplifi, participation signals our commitment to transparent analysis, defensible metrics, and responsible adoption of AI—helping clients move faster with confidence and credibility.