The Complete eDiscovery Solution

The Complete
eDiscovery Solution

The Most Extensive Set of Features on the Market

Servient provides basic features expected in a modern litigation support system plus advanced features required to enable machine learning workflows.

Integrated, Active Machine Learning.

Servient's "active" learning technology is tightly integrated in the application. Servient continuously learns in the background, making it easy for lawyers to harness the power of machine learning.
Learn More

Web-Scale for Matters of any Size.

Servient's architecture is built upon modern "Big Data" technologies including Hadoop and HBase. Servient scales to meet the needs of any size case and enterprise.
See What's Under the Hood

End-To-End eDiscovery



Servient automatically prepares the data for machine learning upon ingestion.

Search and Cull

Search and Cull

Triage collected data while still ensuring maximum recall rate in an easy to use and lightning fast web-scale solution.

Document Review

Document Review

Servient combines the power of our ensemble based machine learning in an easy to use, full-featured review system.



Servient's production module allows for exceedingly fast productions.


Servient automatically prepares the data for the machine-learning eDiscovery workflow upon ingestion.

Statistical Breakdown of Data

Servient’s processing goes well beyond the basic level of indexing and text and metadata extraction that normally happens with a traditional eDiscovery software system. Servient focuses not only on the document but also on the features within the document. During the ingestion process is when Servient breaks down specific textual and metadata based features for use in our machine-learning algorithms. In addition, text from each individual email is compared to the other email to ensure we have a 100% match in our Containment Email process.


Gain Valuable Time

By preparing the data upon ingestion, it saves considerable time later in the process when Servient can have additional machine-learning runs completed in minutes rather than hours or days. This time saving is tremendously important later in the process when deadlines are approaching.

Foreign Languages

Servient’s processing technology has the ability to recognize and process a wide variety of foreign languages. Our system also has the capability to utilize our machine learning features on foreign language data.

Key Functionality Features:

  • File Type Identification
  • De-NIST
  • Text Extraction
  • Metadata Extraction
  • Containment Email Threading
  • Language Detection
  • Sensitive Data Tagging
  • Email Disclaimer and Email Signature Tagging

File Types

While many processing tools struggle with unique file types, Servient has the ability to ingest almost any relevant to eDiscovery file type possible. From AutoCAD files to graphic files such as Photoshop and Illustrator, Servient’s advanced processing tools can handle unique file types without issue.

Web-Scale Power

Servient utilizes a Hadoop set up utilizing HBase and Map Reduce to handle the immense computing power that it takes to complete these processes at lightning speed.

Advanced Tagging Technology

Servient utilizes the newest eDiscovery technology to recognize email disclaimers and email signatures during the data ingestion process. By doing so it allows for more accurate search results and helps the litigation team to get to the data that they are searching for without having to wade through search hits that are unrelated to their search criteria. Additionally Servient’s data processing has the ability during data ingestion to decipher and segregate sensitive information such as social security numbers, patent numbers, and account numbers so that those documents can easily be located within the database.

Search and Cull

Triage collected data while still ensuring maximum recall rate in an easy to use and lightning fast web-scale solution.

Servient greatly improves the rate of recall. Assuming traditional keyword searching was used as a filter for review assignment, the learning model created by Servient’s machine learning can be used to improve the performance of the search. Documents with a high probability of relevance that were missed because of the inherent limitations of basic search technology can be identified.

Instantly Go From Search and Cull to Review

Easily record assignment of documents to review to establish defensibility. Establish rules to streamline the process and apply various de-duplication filters during assignment to review. Duplicates suppressed from review are automatically recorded for production reports.

Immediate Insight of the Litigation Hold Data

Gain immediate insight into the nature of the litigation hold data. Assess and communicate the extent of the requirements at an early stage of the process. Drill into such information as data volumes, file types, data by custodian, data by sources and the extent of duplication.

Powerful Search Technology

Powerful “Search Acceleration Technology” allows a search through terabytes of data to return millions of documents in less than a second. Automated query segmentation splits the query into computationally efficient subparts, delivering incredible search performance over complex queries and provides meaningful statistics.

Email Analytics

Advanced analytics technology automatically identifies alternate email addresses and display names for a selected email address. The searcher can then view samples and combine the various email addresses and display names into a "person object." Queries automatically search for all relevant aliases.

Manage Complex Searches

An easy to use interface can combine over 100 metadata fields with advanced full-text queries into complex searches. A full complement of traditional search methods are available; wildcards, Boolean, proximity, stemming, as well as more advanced retrieval based upon document similarity and relationships. Run multiple searches together on specific queries to target certain custodians, witnesses, yet still understand the total document count being promoted to review.

Visualize Concepts

Concepts are automatically identified and documents are visually organized into “Concept Groupings” and then into sub-topic groups. View and quickly gain an understanding of the documents within the concept groupings that serve as the best starting point in designing the search filter.

Automated guidance in Search Refinement

Text classification technology identifies important keywords. A text classification algorithm creates a score that correlates with the likelihood of the query segment returning a large number of non-responsive documents.

Drill into each query segment and:

  • Test various search refinements to improve the effectiveness of the search.
  • Quickly view a sample of the documents that are returned by the query segment to understand the context of the query in the actual documents.
  • Quickly view all Concept Groups that are related to the query segment to target the search refinement to the relevant concepts in the documents.

Highlighted Search Terms

View all search documents in the litigation hold data set. Search terms are automatically highlighted with in the document for ease of understanding the context of the search terms.

Statistical Sampling Ensures Quality

Based upon the desired confidence measure the searcher can perform statistical sampling to verify the search filter design. Counsel can submit a small sample of documents immediately to review to learn the distribution of responsive/non-responsive within the search results. Counsel can also sample the documents within the litigation hold that have not been promoted to review to ensure that the final search filter is reasonable.

Document Review

Servient combines the power of our ensemble based machine learning in an easy to use, full-featured review system.

Machine Learning And Attorney Review

Servient actively learns from the document decisions made by the legal team during legal review and separates the relevant documents from the irrelevant material. Predictive Review produces meaningful results after the review of only a modest number of documents. As the legal team reviews more documents, Servient continues to learn and further refines the automated document decisions.

Containment Email Threading

Unlock powerful review workflow options and gain up to 40% in reviewer productivity by identifying email within a thread based upon content analysis of each email in the data set.

Servient's analytics identifies the email positioned at the "Top of the Thread" and then verifies that the text of all email tagged as a member of the thread is included within the inline message portion of the Top of the Thread mail.

The reviewer can read the Top of the Thread email and instantly mark the other email in the thread without the need to review each separate email. All email attachments are identified within the thread so that they can be reviewed with the Top of the Thread message. And Servient goes yet further and identifies related threads so that the reviewer can quickly review all similar threads at the same time.

Concept Groupings

Documents are automatically visually organized into Concept Groupings based upon Servient's proprietary text analytics and then into sub-topics. Concept Groups can be used to organize the legal review. Allowing reviewers to consider documents related to the same topic increases the speed of review and quality of the review calls. Concept Groupings also improve the search and analysis of documents. Counsel can visualize the Concept Groups and sub-topics related to the documents returned by any search performed in the Servient Document Review component.

Review Team Management

Servient contains a complete review team management module to handle high volume eDiscovery projects. The review manager automatically generates review batches based upon the rules set by Counsel. The reviewers are then automatically assigned review batches. Detailed real-time statistics regarding the review are only a click away. Information such as individual reviewer productivity and overall status of the project are just two examples of the extensive information available to Counsel to manage the review.

Advanced Near Duplicate Review

Servient's text analytics identifies documents that contain minor differences. Counsel can specify the threshold for the near duplicate algorithm to control the level of differences to be detected. For example, various versions of a contract that have been assigned for review will be detected and presented to the user as a group of "Near Duplicates". Servient automatically displays a comparison of the near duplicate documents, highlighting the differences in the text so that the reviewer can quickly review all near duplicates and make the review call consistent.

Advanced Duplicate Workflow

Servient classifies duplicates to increase the efficiency of the review workflow. Servient identifies binary files and email that are identical and tags them as "Exact Duplicates." The review workflow is built to allow the cascading of review calls to all Exact Duplicates without the need to review each duplicate instance of the document.

Servient also identifies documents that contain the exact same textual content but have some difference in the file metadata. These "Content Duplicates" show the reviewer just the differences in metadata to allow for quick review of like documents. For example, a Word document that has been converted into a PDF will be identified as a Content Duplicate and the reviewer need not look at both files.

As the Document Review component is an integrated portion of Servient On-Demand platform, records of any duplicate suppressed during either the initial data load or the Early Document Assessment phase are recorded so that Counsel can track and report on the duplicates identified at all stages of the eDiscovery process.

Bulk Tagging

Counsel can quickly assign review calls, issue tags and organize documents in subsets with a single click. Any document returned by a search, or any document contained in a concept Group or sub-topic can be selected and bulk tagged. Bulk tagging is the perfect tool for running presumptive privilege filters to assign likely privilege documents for special review treatment.

Audit Tables

Servient maintains an audit table recording the full history of each review call made on the documents. Counsel can use the history of review to cure any anomalies they detect during the manual review or recover from errors identified during the review process.


Servient Document Review allows the reviewer to redact content from documents. Servient automatically generates an image-based document (removing any hidden text layer) and allows the reviewer to "burn in" the redaction box. Security roles control whether a user can toggle back and see the un-redacted version of a document. The reviewer can also delete or edit the redactions from within the application.


Servient provides a full reporting engine thus allowing counsel to generate reports including almost every field in the database. The report engine can quickly generate privilege logs, exhibit lists, etc., to speed case preparation.


Servient’s production module allows for exceedingly fast productions. The production module is a complete, self-service system that allows for granular handling of productions.

Increased Efficiency

The document type can dictate forms of production. Load files can be built to meet the needs of the matter. Once built, templates can be reused for increased efficiency.

Gain Control

Users have full control over the system. Slip sheets, load file formats, formats of doc types are all configurable.

Cloneable Templates

Production formats are governed by customizable and cloneable templates. Standard production templates for U.S. Regulatory Agencies are included.

Complete Production

The module produces in TIFF, NATIVE or PDF and handles stamping and load file creation.