The Department of Defense awarded dMetrics a five-year, $99.5 million production contract in September to advance work on the machine learning platform Minsky, which is designed to rapidly ingest and analyze large volumes of data from diverse sources to identify open source adversarial activity.
Table of Contents
What Is Minsky Platform?
The Defense Innovation Unit said Wednesday Minsky helps DOD analysts create personalized ML agents to screen large datasets to detect and extract actions, entities and relationships relevant to their area of responsibility.
The company initially developed Minsky as a prototype following its selection from a pool of 65 offerors in March 2020 as part of the AI-Based Knowledge Graphing project.
The project was launched in 2019 by the Defense Innovation Unit, the U.S. Army and the Defense Technical Information Center, or DTIC, to advance the development of a customized ML tool to help collect, analyze and produce reports of threat activity from web-based, open-source content.
“DTIC appreciates the efforts of DIU in identifying and helping dMetrics navigate the DoD’s procurement processes,” DTIC Administrator Christopher Thomas.
“The award of this production contract establishes a launch point for DTIC to employ Minsky to expand insight into DoD’s scientific and technical information, allowing our users to visualize connections among researchers, organizations, and topics previously unrecognized,” added Thomas.
Expanding Platform Use Across Federal Government
Agencies within the U.S. intelligence community have also used the platform to glean domain-specific insights from technical literature and combine large, unstructured text datasets to facilitate geopolitical knowledge graphing.
DIU said the dMetrics-built platform could support analysts across the federal government by helping them map criminal network activity, analyze foreign news sources or develop a new search tool.