Labor Market Intelligence & Predictive Analytics to Fast Track Army Technology Leadership

People have always been America’s greatest strength. Today, the Army faces a war for technical talent. To build a vital long-term advantage, we must fill our most consequential ranks. Starting with commissioned officers, we must identify technology leaders who can maximize the effectiveness and deployment of emerging technologies -- by acting as the linchpin between the operational and technical community. Namely, Emerging Technology Leaders (ETLs). At a conference in 2019 by the Association for the United States Army (AUSA), Gen. McConville, the Chief of Staff, told the audience how critical skills are currently hidden in the Army’s personnel system. He found out that during the Afghanistan surge, he had a supply sergeant who ran an engineering design firm -- who as it turns out, designed all the forward operating bases in Afghanistan. And a major who worked for the Texas Highway Department, who helped the troop surge build highways. Yet none of this vital information was visible in the existing personnel system. The Chief had to manually ask people to fill in an Excel spreadsheet to discover their vital skills. This highlights the importance and great potential of the modernization of ETL talent management -- in particular, making ETLs visible. And with a modern data science approach, we can go beyond Excel spreadsheets and incorporate not just hard skills, but also critical human factors.The first step: identify ETL talent by mapping hard and soft skills with the right behavior attributes. At Voltera, our management is composed of serial technology entrepreneurs who have mastered this art in a systematized fashion -- for example, our CEO is an Army-connected disabled veteran who has led the commercialization of multiple emerging technologies. We will apply modern data science techniques such as k-means clustering -- which our Principal Investigator used successfully at Qualcomm to classify 15,000 hardware engineers and subsequently optimize their engineering workforce, including at multi-million dollar projects -- to identify ETLs from input data. To capture the skills needed for the progression to become an ETL, we will describe the clustering models that could be used to segment the augmented input data to show and/or infer the hard skills, soft skills and behavior attributes for ETL identification. We will also consider supervised learning models such as logistic regression, k-nearest neighbors, decision tree-based classifiers or neural network classifiers. Our data partner Emsi is the world leader in labor market data and analysis tools and holds the country’s preeminent databank of resumes, job postings, skills classifications and natural language processing (NPL) resume parsing tools. It provides us the ideal warehouse of data to train the envisioned software and revolutionize ETL talent management in the Army.