Executive Q&A: Developing Data-Focused Professionals
Employers continue to face challenges finding and hiring the talent they need. Amarda Shehu, Professor of Computer Science, Associate Dean for AI Innovation, and AVPR of the Institute for Digital Innovation at George Mason University offers insights about developing that talent, especially the role higher education and government can play.
- By Upside Staff
- September 18, 2023
Upside: What are the data-focused jobs enterprises are having the toughest time finding employees to fill?
Amarda Shehu: It is worth reminding ourselves that the data revolution has redefined entire enterprises. Nowadays, data skills are in high demand in almost every industry. The top industries for data analytics skills are quite diverse, including banking and securities, insurance, healthcare, education, media and entertainment, transportation, manufacturing, and even government.
There is a shortage in general of employees with data science, data engineering, or data analytics skills. Distinctions are often made based on the position requirements and the skills employees bring with them. If you search for data science jobs, you will find a variety of positions, including data analyst, data engineer, database administrator, machine learning engineer, data scientist, data architect, statistician, business analyst, data and analytics manager, and more.
According to CIO.com’s 2023 State of the CIO survey, the top three roles enterprises are finding a hard time filling are cybersecurity (22%), data science/analytics (22%), and AI/machine learning (20%).
It is encouraging that universities have started to be more responsive to enterprise needs in the past few years through new concentrations, certificates, and even new bachelor’s programs specifically in data science, cybersecurity, and other fields at the intersection of data, AI, and security. However, we are not graduating enough students to power these enterprises.
I think that this is the main challenge for enterprises; there are just not enough students coming out of college with the skills they need. It is ultimately a numbers game. The numbers get worse if an enterprise needs highly sophisticated or specific skills, say in machine learning, predictive analytics, or generative AI (think chatbots and large language models such as ChatGPT here), so many enterprises have removed the requirement for a formal four-year education and have taken to training employees in house.
Are enterprises generally successful in developing these skills internally? Why or why not?
Amazon has been successful at offering on-the-job learning opportunities and apprenticeships. However, Amazon trains employees specifically for its own needs. This focus is partly why it has been successful. As the second largest enterprise, it also has the capacity needed to support multiple training pathways, such as classes online, in person at a local campus, or on site in any of the more than 110 classrooms in fulfillment centers in 37 states.
Many smaller enterprises do not have this capacity, so it is crucial for them to connect with universities and use universities as the primary infrastructure for “reskilling” and “upskilling” current and future employees. In fact, I could argue that this should be at the forefront of the mission of every institute of higher education.
For example, my employer, George Mason University (GMU), has created several concentrations and programs (at the undergraduate, master’s, and Ph.D. level) in response to employer needs, in close collaboration with them, as well as proactively anticipating future trends and technologies. GMU is using its centers and institutes to forecast future technologies and assist the various colleges in creating nontraditional, fast-tracked programs for workforce development and training. These include skill development in emerging technologies at various levels, from high-school students to community-college students, all the way to GMU students and employees. Several executive development programs at GMU offer skills in data analytics to employees seeking data-focused jobs.
What role can higher education and/or government programs play in developing employees to meet enterprises’ needs?
Higher education should be at the forefront of developing employees to meet enterprise needs. In fact, one of the core functions, central to the mission of a higher education institution should be to drive economic development. I believe that the current fast-moving technological landscape requires universities to be responsive and visionary and train students for the jobs of the future.
Many universities have been caught unprepared for the exploding demands in AI skills. Most educational programs are traditional (four years) and do not necessarily give students the specialized in-time skills they need for these jobs. Deloitte had an interesting article about “AI whisperers” as the job of the future, referring to enterprises’ need for employees who deeply understand machine learning algorithms, data structures, and programming languages. Such jobs are already being advertised. An institute of higher education needs to be agile enough to create concentrations and certificates that quickly provide students and existing employees with just-in-time skills.
Government should support these efforts. Agencies should provide funding to incentivize four-year institutions (and community colleges) to implement diverse programs to respond to current and future needs of industry; the National Science Foundation’s Directorate for Technology, Innovation, and Partnerships is a good example of a funding agency trying to do that, but universities need more. They also need specific support to diversify their student and employee pool. We need something like what the GI bill did for World War II veterans. We also need specific programs to include students who dropped out of higher education programs.
Several reports have been written on how to bring these students back. Millions of these students can complete their degrees. According to some recent data from a ReUp Education Report, it is estimated that 89% of students who have left school with “some college, no degree” express interest in returning. Universities need to attract these forgotten students and support them to complete their education by providing flexible pathways so they can participate in our digital economy, paying particular attention to minority students.
What types of higher-education/government programs work? What programs are less successful?
I hinted above at just-in-time skills and just-in-time programs. Typically, universities move slowly. There is inertia, and you can argue it is by design: universities are most comfortable with a traditional four-year education. They know how to do that, and education boards that approve these programs are also comfortable with that format. However, a four-year education does not speak to all students or speak to their needs and where they are in life. I would argue that by only focusing on four-year programs we are disincentivizing many forgotten students and minorities and are not serving the fast-moving technology needs of enterprises.
I love what GMU is doing here through its continuing professional education programs and the Mason Virginia Promise: a pathway to a bachelor’s degree or help starting a business for every Virginian who wants one. GMU’s Innovation and Economic Development Office, headed by Paula Sorrell, has been highly effective at keeping this promise and training the next entrepreneurs in Virginia. I also mentioned earlier that institutes are great infrastructures to support skills development. For instance, the Institute for Digital Innovation, which I lead with Kammy Sanghera, has many such programs for high school students, undergraduate students, and employees, and we are actively growing these into a mature infrastructure for just-in-time training. We need more enterprises to partner with us.
How do you see enterprises’ needs changing over the next two to three years?
If you open any news site, almost every day there is some article on AI. I may be biased because I have long been an AI researcher and continue to be. In my additional role as the Associate Dean for AI Innovation at the College of Engineering and Computing, I am a firm believer that AI skills will be essential for white collar workers and our digital economy. Already enterprises are looking for AI professionals. We need to innovate our education and training programs to prepare employees with these skills and address the current anxiety among white collar workers.
I was at a conference for marketing professionals a few months ago, and I presented opportunities for how they could use chatbots to improve their productivity. The reception was great, and many attendees were excited. However, some were anxious and unsure about their ability to acquire these skills on their own. They need universities to lead and develop programs that make it easy for them to get these skills while shuffling a full-time job and family obligations. GMU is already responding and planning expansion of continuing professional education programs.
Can higher-ed/government programs keep up with this demand? If not, how do you envision them changing?
I don’t think it is a question of can. We just need to, so the real question is how. If we want the U.S. to be competitive on the world stage, we better keep up with the demand. Flexibility is key. Willingness to try new programs and not fear failure is important here. Government support for these efforts is essential for institutes of higher education to be bold and, dare I say, even experimental and develop effective infrastructures for just-in-time skills training.
Personally, I am excited about the future, and I am particularly energized and thrilled about how GMU is expanding its portfolio of workforce development and training programs. GMU is doing some great things. All I can say is “Stay tuned!”