Nuix and Serco NA Partner to Score Prizewinning Result in US Navy AI Automation Challenge

Global software company Nuix announced that alongside partner, Serco Inc, that it was a winner in the US Navy’s Controlled Unclassified Information (CUI) Automation Challenge sponsored by the US Navy Office of Naval Research. In an increasingly tense geopolitical environment, managing the various types and volume of classified information correctly protects the integrity and confidentiality of the information they contain. The US Navy’s recent “CUI Tool Automation Challenge” highlights the need by complex enterprises to machine automate the document assessment using recent advances in Artificial Intelligence (AI). The Nuix/Serco team were one of three participants to emerge from this challenge as p**** winners.

“We were delighted to work with Serco on this challenge,” says Oliver Harvey, Executive Vice-President, Customer Strategy & Innovation for Nuix. “Working together we were showcasing our capabilities to an important audience – using an MIT-award-winning AI to process vast volumes of important contextual information and successfully automating a laborious process open to human error.”

Recommended AI: UTB Bot Unveils a New Way to Leverage Automation and Cryptocurrencies

Correctly setting the classification level of US Government and Department of Defense documents allows DOD organizations to reduce the barriers to effective – and vital – information-sharing with partners and allies. But once information is assigned a classification level, it’s a complex and labor-intensive challenge to maintain the chain of trust as various human beings interact with it.  A problem made even more complex when the number of documents that must be correctly handled goes into the millions. To address this challenge, the US Navy’s Naval Surface Warfare Centre launched a competition to find solutions providers with viable approaches to automate the classified document assessment for a particular type of non-classified information called controlled unclassified information (abbreviated CUI).

While the CUI challenge was designed specifically for the US Navy and the Defense establishment, it’s directly and increasingly applicable to a wide range of commercial and governmental organizations today. In fact, the challenge is ubiquitous: How to leverage emerging technologies (like AI/ML/NLP) to automate and streamline the ingestion, interpretation, classification, and prioritization of a vast and steadily increasing volume of information to drive better decisions and support business objectives.

Recommended AI: AI Smart Chain Ecosystem Launches, Bringing Artificial Intelligence to Crypto Space

To capture the top-three finish, the Nuix/Serco team had to formally present a solution design, participate in a rigorous multi-month down-selection process, and ultimately deliver a self-installing prototype software product, which the US Navy tested in their lab. A combined team of Nuix Inc and Serco Inc subject matter experts worked collaboratively using a small data set to train the Nuix NLP product to automatically detect CUI in unstructured text documents, provide a confidence score for each document, and bundle and deploy the solution for ultimate deployment on a US Navy lab network. Once trained and deployed, the prototype successfully flagged CUI content leveraging key Nuix NLP features like a proprietary large language model, text categorization, document-type classification, context-aware extractions, and Risk Engine.

“Serco’s world-renowned services and deep domain expertise, coupled with Nuix’ ground-breaking AI technology, made us a formidable competitor for this complex text analysis challenge,” says Bob Trabucchi, Senior Serco AI/ML Solutions Architect. “Our success on this project is even more impressive when you consider that it was achieved by just two engineers on a part-time basis. And we’re just starting to scratch the surface of what’s possible.”

Recommended AI: Cognni Launches AI-Powered Automated Infosec Risk Assessment Product

[To share your insights with us, please write to sghosh@martechseries.com]