Product Advantages

Inference has combined the framework and advanced analytics of the Autonomy Intelligent Data Operating Layer™ (IDOL™) engine with accepted Boolean (i.e., keyword) searching, allowing users to follow standard procedures and established practices of legal review, while also offering advanced features that highlight and recommend documents based on similar concepts. Inference’s configurable ASP hosting solution provides users and system administrators with the ability to enable or disable features as needed. Moreover, unlike other systems, Inference does not automatically code documents but provides users with an intuitive and organized method of review.

Key Product Features

Advanced Bayesian Meaning-Based Analytics
Leveraging industry-leading technology established at Cambridge University, Inference utilizes Autonomy’s “meaning-based computing” algorithms to power its legal feature set. This powerful searching technology can be deployed for early assessment, meet & confer preparation and linear review prioritization. Inference ingests volumes of unstructured data and enables search, retrieval and content analytics, allowing the reviewer to determine relevance. Meaning-based computing extracts the digital essence of a document, allowing it to be clustered and hyperlinked to a range of related evidence. Autonomy’s advanced pattern-matching technology utilizes Bayesian Inference and Claude Shannon’s Principles of Information, which treats information as a quantifiable value to produce these advanced results.

Accelerated Review
Inference analyzes and streamlines user workflow to dynamically accelerate the review process. Using Inference concept and clustering tools, users can quickly segment important documents from potentially non-responsive, low-priority information. The entire database can then be assigned to appropriate levels of reviewers, who can use advanced “document reveal” and robust review and production tools to quickly execute the review process. Use of analytics provides a significant ROI due to review time acceleration when compared to standard linear review software.

Intuitive, Attorney-Friendly Interface & Base Tools
Inference has merged the advanced features of the Autonomy analytics technology with the Boolean searching necessary for the stringent processes and standards applicable for legal review and production. The attorney-familiar query syntax is integrated with a simple user interface, enabling attorneys to control the datasets using accepted standard Boolean searching, which can then be processed through the analytics stage in real time. This allows attorneys the control to review and “ask” what’s inside the various datasets they determine are relevant. Further, it provides a familiar look, feel and logical methodology trusted by attorneys.

Trusted Autonomy Technology
The advanced analytics engine of Inference is powered by the proven Autonomy IDOL™ engine. Deployed to thousands of corporations and organizations worldwide, Autonomy's technology has been selected by both the U.S. Securities and Exchange Commission and the New York Stock Exchange for the purpose of performing regulatory investigations. These selections not only provide unmatched levels of acceptance and approval of the technology, but also a unique opportunity for ensuring that attorneys can support the use of the advanced searching and analytics to provide proper defenses for their clients.

Unlimited Scalability and Search Speed
A key differentiator of Inference is its ability to scale. By having the data in a distributed index environment, we do not experience the size or speed limitations of other analytics and review providers. As other products utilize standard databases (e.g., SQL), each experiences limitations based on a linear architecture. As the size of databases increases, performance is inversely affected. Inference is built on an architecture that allows the system to rapidly scale beyond the requirements of even the largest litigation cases. As the diagram here shows, the system incorporates a Distributed Action Handler (DAH). The DAH translates queries and index instructions from Inference to the Autonomy IDOL™ index engine. As data increases, additional engines can be introduced to continually scale while ensuring peak performance.

Ability to Analyze Data without Proprietary Processing
Inference is the only user-driven analytics software that does not require proprietary processing to index the data and enable the use of analytics. By providing Inference with a load file generated through standard EDD processing, the system can upload the data and perform the necessary mathematics and analytics during the ingestion process.

Broad Array of Data Formats
Utilizing the Autonomy IDOL™ engine, the Inference software is language independent, learning enabled and format agnostic. This allows for the analysis of unstructured (HTML, word processing, rich media), semi-structured (XML) and structured (Oracle, Lotus Notes) data. Data can be automatically categorized, tagged, linked, retrieved and profiled in real time regardless of repository. Future releases will leverage other Autonomy components, which enable the system to encompass multimedia data as well.

Configurable Workflow
Inference offers a highly configurable workflow and feature set that enables case administration and security flexibility. Features can be disabled for certain reviewer levels, while enhanced analytics provide a method to distribute prioritized review to select individuals.