NLP ONE TWO THREE: Intuitive DoD Dictionary, Patterns, and Forecasts

The complexity of acronyms, jargon, and terms in the Department of Defense is well-suited to an NLP solution. However, we understand the significant problems that can arise when introducing sophisticated and scattershot machine learning techniques to more traditional executives and leadership. During this Phase I, we propose to leverage our product management and engineering talent to deliver a white paper, plan, and rapid prototype for DHRA. These materials will fully explore the problem and solution space, and provide a clear-eyed vision and plan for Phase II and beyond. Our focus will be to leverage best-of-breed open source technologies to both future-proof this project and avoid "vendor lock-in." At the same time, as practitioners of a powerful mix between "waterfall" and "agile" development, we believe that the best solutions come from exploring the solution space with actual working software, even if very rudimentary. As such, we aim to use publicly available lookalike data (e.g., public DoD websites or press releases) to immediately develop a working prototype that can drive forward customer discovery and provide a powerful means to "kick the tires."