Asset 1
Digital Reasoning Sets New Standard in Real-time Email Analytics for Risk and Compliance
Machine Learning Platform Shatters Previous Quality and Performance Benchmarks

NASHVILLE, Tenn. July 29, 2013 – Digital Reasoning, the company that enables automated understanding of human communication, today announced the General Availability of Synthesys® v3.7, the latest release of their patented machine learning platform. As a result, organizations can now monitor and analyze email traffic in near real-time to quickly identify risks and suspicious behavior that could present regulatory or reputational risks.

“Big data analytics in fraud and security is fast becoming a big business and ‘must have’ technology for organizations that generate massive amounts of data and, at the same time, need to understand offenses and infractions against their enterprise that don’t necessarily show up in security and fraud monitoring systems,” noted Avivah Litan, vice president and distinguished analyst at Gartner[1]. “Big data analytics allows an organization to organize and analyze vast amounts of structured and unstructured information to facilitate the detection of rogue employees, partners, or criminal or collusive rings of fraudulent or abusive activity.”

Synthesys is an advanced Machine Learning platform built to analyze human-generated data and transform it into a rich knowledge base. Synthesys is designed to take advantage of distributed computation technologies such as Apache Hadoop and Storm; and provides distributed data storage and retrieval on multiple technologies such as Cassandra, HBase, or Accumulo. With the release of Synthesys v3.7 the following enhancements are immediately available to customers:

Real-time Ingestion: Synthesys now includes a streaming ingestion architecture based on the Storm open-source system. The resulting analytics are persisted directly into the Synthesys knowledge base.

Email Analytics: Synthesys can now resolve references to people, organizations and locations across e-mail headers, salutations, signature blocks, and e-mail body. It is able to detect complex events within the e-mail text and relate them back to resolved concepts. In other words, Synthesys doesn’t just find names of people, places and organizations in e-mails; it determines how they are connected and what they mean.

Entity Resolution: Synthesys delivers a dramatic improvement resolving specific mentions of the same entity into a global concept. For users, that means finding more, richer concepts in less time.  Synthesys can now resolve 10X the number of mentions resolved in 25% of the time, setting a new standard for entity resolution across human-generated data.

Region Classification: Synthesys can now classify specific regions within a single document into different categories by training a document classifier to specify features within a region of a document for inclusion in the model being trained. When performing document classification, different categories will be assigned to different regions within each document allowing users to more quickly analyzing the predictive coding decisions for relevance related to their particular issue.

“Synthesys v3.7 provides breakthrough analytics for organizations who are struggling to meet their oversight requirements for electronic communications,” said Bill DiPietro, vice president product management at Digital Reasoning. “We are transforming the way compliance teams can connect the dots and identify risks by helping them understand what people are talking about, to whom and why in real-time.”

Additional information:

– For a free copy of the latest copy of the Synthesys Technology Overview go to

[1] Gartner “Use Big Data Analytics to Solve Fraud and Security Problems”, Avivah Litan, 29 March 2013