Upon graduation, recent applied mathematics Ph.D.s are routinely confronted with what often feels like a career-defining choice: academia or industry? Some choose to pursue an industry-based career at a company, government institution, or nonprofit, while others work towards a tenure-track position at a university. At the 2017 SIAM Annual Meeting, held in Pittsburgh, Pa., this July, a professional development session entitled “Beyond Research Training: Critical Skills for Careers in Industry” offered attendees advice for becoming more competitive in industrial positions. Drawing from their own professional experiences, the five panelists shared examples of industrial careers and proposed tips on valuable skills, industry trends, and opportunities for growth.
Employment at U.S. national laboratories is a common non-academic path, with positions for both postdoctoral researchers and early and mid-career level professionals. Lois Curfman McInnes (Argonne National Laboratory) and Carrie Manore (Los Alamos National Laboratory) espoused the versatility of research at laboratories. “It’s a really fun, interdisciplinary environment for applied math and computational research,” McInnes said.
Manore, who focuses primarily on disease modeling, echoed this thought, adding that the national labs work on a remarkably wide variety of projects in chemistry, biology, and other fields. “Don’t count yourself out of the national labs because you don’t feel like you fit exactly with what they’re doing,” she said. “If you have the math and programming skills, they’re open to you switching directions. And once you get there, there are a lot of opportunities to do so.”
Manore encouraged attendees to apply for summer workshops and programs at the labs, which match students with potential future mentors. “It’s a good way to get experience that could apply to other labs, but also for the lab to get a feel as to whether it’s a good fit,” she said. Labs are often willing to take a chance on students and postdoctoral fellows because they are fairly inexpensive. Like other facets of industry, labs are characteristically focused on deliverables, and put a considerable emphasis on applicants’ published papers or self-written software.
Recently retired from AstraZeneca, Jeff Saltzman spent 18 years at Los Alamos and 13 years at Merck, where he developed an applied mathematics group. He provided input on valuable resources for a career in pharmaceuticals research, such as mathematical biology courses, and cautioned that job-seekers should be careful with titles. “If you’re looking for jobs in the pharmaceutical industry, you’ll rarely see a job labeled ‘applied mathematician,’” Saltzman said. “Yet there’s a rich array of work that is applied. When looking at job boards, look beyond jobs labeled ‘applied mathematics.’” Instead, he suggested searching for jobs requiring biology or math experience.
Unsurprisingly, attendees were particularly interested in the differences between academic and industry-based positions. Though academia is normally thought to offer more freedom, Saltzman indicated that the two are not necessarily as dissimilar as they appear. “Industry seems highly constrained, like you have to work on a specific project,” he said. “But you end up steering the work to the area that you want to work on.” Because there is less pressure to publish in industry, Saltzman finds the publications to be more meaningful. While transitioning between the two settings is possible, he suggests establishing a personal timeline and deciding in advance how many—and what types—of publications would be necessary for a switch. “It’s not going to be a fast transition,” he warned. “But I know a lot of people who transitioned from industry and academics and they were very deliberate about it. Many academic departments are looking for people with some industry experience so their students can learn something practical.”
Henry Warchall initially planned to spend only two years in a temporary rotator position as a program director at the National Science Foundation (NSF), but enjoyed the work so much that he stayed on. “The Division of Mathematical Sciences at the NSF is actively promoting the idea that graduate students in mathematics particularly, and statistics as well, should be exposed to at least the possibility that they could pursue a non-academic career,” he said. “When you’re thinking of applying for jobs in industry, you should bear in mind that you have something rare. You have the ability to approach an open-ended problem and potentially solve it.”
Panelists encouraged the audience to dabble in software, as industry employers look favorably upon programming experience. “There’s a lot of work done with interpretive language these days,” Saltzman said, citing Python, R, MATLAB, and Mathematica as frequently-used programming languages. He urged students to have patience, be flexible, and experiment with different software. “It’s not so hard to learn R if you’ve learned Python,” he said. “In the end, it’s all about finding the right code with which you can share and communicate your thoughts.”
Efficient use of software is particularly useful in collaborative industrial settings, which bridge diverse fields increasingly linked by big data. “I believe software is really a foundational part of science and engineering as a means to collaborate across interdisciplinary boundaries and sustain those collaborations,” McInnes said.
One trend in the discipline replete with opportunities is big data. Saltzman touched on the changing nature of high-performance computing, given the explosion of data in nearly every field. “What you’re seeing in companies is cloud services like Amazon,” he said. “There’s a huge demand for people to process information, especially genomics and medical information, at a level that demands them to really understand.”
Kristin Bennett (Rensselaer Polytechnic Institute) reiterated Saltzman’s sentiments about data processing and spoke of the value mathematicians bring to the new age of big data. “Industry wants you, pharma wants you, everybody wants you,” she said. “Virtually we are collecting data in every industry — for finance, for drug discovery, for government, for roads. Data [utilization] is a fundamental skill that we all need to use in the future.”
Data processing has applications in healthcare, disease mediation, finance, portfolio and investment management, and marketing, in addition to computer science and machine learning. Internships and workshops offer students opportunities to work with data, and a significant portion of math industry workshops are data-focused. “Right now, there’s a huge shortage of people who can do data analytics,” Bennett said. “I’m encouraging you to get even a little bit of data experience. This can be a differentiating factor to get your foot in the door with companies and open up opportunities in the future. If you can do data, you can do anything.”
Communication is also a critical skill for mathematicians in industry. Warchall emphasized the importance of understanding your audience, both when interviewing and during employment. “When you’re interviewing for a job, you have to prepare to talk to people who are not in your field,” he said. “They might not even be mathematicians. I’ve seen a lot of certifiably brilliant mathematicians who have failed miserably in public talks.” Warchall then introduced a helpful analogy. “Pretend it’s like teaching,” he said. “Think about what students know and don’t know, and plan your lessons from there. Mathematicians tend to assume an awful lot about their audience, and you really can’t. There are lots of things that people don’t know that you know really well.”
Once securing a position, strong communication skills remain just as imperative. “Often times when you’re working in science projects in Department of Energy labs, you’ll be working with people who have complementary skills or expertise,” McInnes said. While learning each other’s strengths and developing joint skills takes time, it is undoubtedly rewarding — especially because industry and business jobs generally emphasize interdisciplinary work. Each discipline also commonly has a slightly different communication technique, which Manore admitted can be off-putting at first. “Be prepared for those differences in communication styles and try not to be offended,” she said.
Upon establishing oneself in a given area, there are naturally opportunities for growth. Upward mobility in industry is typically first lateral, in that one moves to a different job before directly advancing. In some cases, industry Ph.D.s will ultimately find themselves in managerial positions. “I think it’s almost impossible to avoid management in any of these fields,” Warchall said. “That’s not true without exception, but I think the expectation is that you’ll eventually turn into a leader.” Saltzman clarified that he encountered different levels of leadership during his time at Los Alamos, given practical constraints in the number of managers an organization can have. “I thought everything was management in industry,” he said. “But it turns out there are parallel tracks as well. For every manager, you should have 10 people who are doing work.”
The panel concluded with a brief discussion on funding. “I think there’s an illusion that working for the government, it’s easier to get funding than working for industry,” Saltzman observed. “It’s not. If you think you’re going to make a decision based on work environment, the variation within any particular area is much greater than across.” When asked about industry job security given the lack of tenure, Bennett had advice pertinent to almost any career field. “You should go for the position of what interests you,” she said. “Not what you think will be most secure.”
|| Lina Sorg is the associate editor of SIAM News.