Daegis, a provider of e-discovery software and services, this week announced the launch of its Hadoop-based technology assisted review (TAR) solution, Acumen. The new offering is designed to reduce the high costs associated with document review, while increasing accuracy and defensibility.

Fully integrated into the Daegis eDiscovery Platform, Acumen reduces the number of documents for first-pass review by 70 to 90 percent, cutting first pass review costs by up to 50 percent and substantially increasing the speed and efficiency of the review process.

Unlike e-Discovery technologies that focus solely on black-box predictive coding, Acumen implements a workflow and process-driven approach based on the latest big data advancements, innovative software algorithms and proven best practices in human review. Daegis' focus on scalability future-proofs corporate concerns around data growth and the challenge of managing, storing, and securing data as well as sorting and identifying the relevant pieces to meet e-Discovery obligations.

Acumen utilizes Hadoop-based distributed computing with a modified form of Probabilistic Latent Semantic indexing to rapidly create and ensure highly accurate data maps that represent review document sets. The Acumen solution is seamlessly integrated into the Daegis review platform to leverage existing advanced search capabilities and process-driven human review.

Important elements of Acumen include:
  • Transparency. The Acumen dashboard offers users in-depth insight into review progress, accuracy, cost metrics, and the process and data itself.
  • Defensibility without Delays. Acumen has built-in continuous learning and checks and balances designed to increase the defensibility of the production set. For example, if new data is added to an existing population, Acumen automatically resamples and adjusts projections in real-time based on the entire new population without requiring a complete restart of the sampling process.
  • Advanced Iterative Learning. Acumen's iterative learning is designed to seamlessly integrate technology into well-established document review processes, thereby maximizing human review investment. High resolution data maps are used to train reviewers, who in turn use the maps to train the system to identify responsive documents. The system is designed to identify and push ambiguous documents to the review team, thus maximizing the training benefit of each reviewed document for both human reviewers and the technology.
  • Scalability. Because it is built on the Hadoop architecture, Acumen expands and contracts computing capacity based on current need, enabling streamlined integration of additional data as the matter progresses.
  • Seamless Platform Integration. Acumen is fully integrated into the Daegis e-Discovery Platform enabling users to easily make use of TAR functionality across all current and future matters. Acumen is also available as a stand-alone review solution

The Acumen solution is free of charge for clients using the Daegis e-Discovery Platform for processing and hosting. Acumen is also offered as a stand-alone solution with pricing available per gigabyte or as a bundle with other services.