Ox: Analytical and decision support tool enabling human-machine teaming

Ox Intel is removing barriers to adoption of AI in human-led analysis and decision-making. As analysts and decision-makers suffer from information overload, there is substantial interest in leveraging next-generation technology such as ML to augment human capacities. However, traditional human-led decision-making and analytic processes have presented barriers to adoption of next-generation technology. Current decision-making and analytical processes are unstructured, inconsistent, and opaque; they waste time and energy, allocating tasks to people that could be performed more efficiently and accurately by machines; and they fail to collect essential data around past judgments to improve future analysis. This creates a major opportunity cost and competitive disadvantage, both for the Defense community and other sectors. Ox bolsters the analytical judgment-making process—an essential area entirely underserved by available technology, and an area where crucial human qualities and capabilities are at risk of being displaced by “black box” automated processes.  Ox is a web-based SaaS platform that leverages structured analytic techniques to structure data around human judgments, enforcing rigor, consistency, and transparency. In the Ox platform, users capture their analytical and decision-making criteria for a range of use cases. Ox then provides a consistent and transparent scoring mechanism for analytic confidence, and enables users to track the performance of their assessments. Ox helps users leverage data to refine their analytical frameworks over time and to scale proven frameworks across the enterprise. Ox also helps identify future opportunities for automation that can relieve overburdened analysts suffering from data overload while preserving the essential qualities of human judgment. The resulting output optimizes human-led analysis and helps train ML algorithms to improve intelligent decision-making at scale. The Ox tool is designed for users with no modeling or data science skills, building on existing MCDA methodologies in an intuitive, interactive interface. Ox is easy to learn, easy to teach, and requires no special equipment or software. It can be used alone or collaboratively, and used quickly to meet tight deadlines. Ox works on a range of analytical and decision-making challenges, both within and outside of the Defense ecosystem.  With Phase II funding and AFWERX support, Ox will help USAF analysts and other decision-makers make high-quality judgments, faster, by powering proven analytic techniques with AI. To meet this goal, Ox’s Phase II effort will accelerate development of the web-based Ox platform, working with Analysis & Exploitation Teams (AETs) within the 480th Intelligence, Surveillance, and Reconnaissance Wing. Our Phase II effort will ensure that AETs can leverage Ox’s customizable frameworks to conduct structured, rigorous, and consistent analysis and integrate AI and ML models directly into their workflows.